{ "name": "E2B alternatives, compared against E2B", "url": "https://e2balternatives.com/", "license": "CC BY 4.0", "cite": "E2B Alternatives, https://e2balternatives.com/", "updated": "2026-11-11", "fields": [ { "key": "isolation", "label": "Isolation", "type": "enum", "description": "Isolation technology the vendor states for each sandbox.", "filterable": true, "sortable": false, "enumValues": [ "Firecracker microVM", "Cloud Hypervisor microVM", "microVM", "VM", "gVisor", "container", "Hyper-V", "WebAssembly", "pluggable (gVisor/Kata)", "process sandbox", "restricted runtime" ] }, { "key": "isolation_options", "label": "Other isolation options", "type": "string", "description": "Alternative runtimes the vendor offers besides the default (e.g. VM runtime, gVisor for GPU).", "filterable": false, "sortable": false }, { "key": "deployment", "label": "Deployment", "type": "enum", "description": "Managed cloud service, self-hosted software, or both.", "filterable": true, "sortable": false, "enumValues": [ "managed", "self-hosted", "both" ] }, { "key": "billing_basis", "label": "Billing basis", "type": "enum", "description": "What the meter counts: wall-clock allocated resources, active CPU only, requests/plan quota, session-hours, or free software.", "filterable": true, "sortable": false, "enumValues": [ "wall clock", "active CPU", "memory while active", "plan quota", "session-hour", "bundled units", "free (OSS)" ] }, { "key": "price_vcpu_hour", "label": "vCPU price", "type": "money", "unit": "USD/vCPU-hr", "description": "Published on-demand CPU rate normalised to USD per vCPU-hour (per-second rates multiplied by 3,600; per-core rates divided by vCPUs per core where the vendor defines it). Lowest public tier, default region. 0 means the vendor states no separate CPU charge.", "filterable": true, "sortable": true }, { "key": "price_gib_hour", "label": "Memory price", "type": "money", "unit": "USD/GiB-hr", "description": "Published memory rate normalised to USD per GiB-hour (GB treated as GiB where the vendor does not distinguish). 0 means the vendor states memory is free.", "filterable": true, "sortable": true }, { "key": "price_sandbox_hour_from", "label": "Sandbox from", "type": "money", "unit": "USD/sandbox-hr", "description": "Smallest published per-sandbox hourly price, where the vendor publishes one.", "filterable": false, "sortable": true }, { "key": "paid_plan_from_usd_month", "label": "Paid plan from", "type": "money", "unit": "USD/mo", "description": "Cheapest paid platform plan fee (before usage).", "filterable": false, "sortable": true }, { "key": "free_tier", "label": "Free tier", "type": "bool", "description": "A free plan, free monthly allowance or sign-up credit exists.", "filterable": true, "sortable": false }, { "key": "free_tier_detail", "label": "Free tier detail", "type": "string", "description": "What the free tier or credit includes, as published.", "filterable": false, "sortable": false }, { "key": "cold_start_ms", "label": "Cold start", "type": "number", "unit": "ms", "description": "Vendor-published create/boot time in milliseconds (a vendor claim, not our benchmark). Only filled when the vendor publishes a number; see startup_claim for percentile and conditions.", "filterable": false, "sortable": true }, { "key": "resume_ms", "label": "Resume time", "type": "number", "unit": "ms", "description": "Vendor-published resume-from-pause/standby/snapshot time in milliseconds.", "filterable": false, "sortable": true }, { "key": "startup_claim", "label": "Startup claim", "type": "string", "description": "The vendor's startup/resume wording with conditions, quoted or closely paraphrased.", "filterable": false, "sortable": false }, { "key": "max_session", "label": "Max session length", "type": "string", "description": "Longest single sandbox session by plan, as published.", "filterable": false, "sortable": false }, { "key": "persistence", "label": "Persistence / snapshots", "type": "bool", "description": "Filesystem state can be kept between runs (snapshots, persistent disks, volumes).", "filterable": true, "sortable": false }, { "key": "pause_resume", "label": "Pause / resume", "type": "bool", "description": "A sandbox can be paused, suspended, hibernated or put in standby and resumed later.", "filterable": true, "sortable": false }, { "key": "memory_snapshot", "label": "Memory state kept", "type": "bool", "description": "Pause/hibernate preserves running memory state, not only the filesystem.", "filterable": true, "sortable": false }, { "key": "fork", "label": "Fork / branch", "type": "bool", "description": "A running or snapshotted sandbox can be cloned into new sandboxes.", "filterable": true, "sortable": false }, { "key": "gpu_support", "label": "GPU", "type": "bool", "description": "GPU sandboxes are offered.", "filterable": true, "sortable": false }, { "key": "self_host", "label": "Self-host", "type": "bool", "description": "Can run on your own hardware or cluster (open source or licensed).", "filterable": true, "sortable": false }, { "key": "byoc", "label": "BYOC / VPC", "type": "bool", "description": "Vendor-managed deployment inside your own cloud account.", "filterable": true, "sortable": false }, { "key": "open_source", "label": "Open source", "type": "bool", "description": "Core sandbox runtime is published under an open-source licence.", "filterable": true, "sortable": false }, { "key": "license", "label": "Licence", "type": "string", "description": "Open-source licence of the runtime, if any.", "filterable": true, "sortable": false }, { "key": "sdks", "label": "SDKs / languages", "type": "string", "description": "Official SDK languages (or supported execution languages where noted).", "filterable": true, "sortable": false }, { "key": "regions", "label": "Regions", "type": "string", "description": "Published regions or region count.", "filterable": false, "sortable": false }, { "key": "max_vcpu", "label": "Max vCPU per sandbox", "type": "number", "unit": "vCPU", "description": "Largest published vCPU count per sandbox on any self-serve or enterprise plan.", "filterable": false, "sortable": true }, { "key": "max_memory_gib", "label": "Max memory per sandbox", "type": "number", "unit": "GiB", "description": "Largest published memory per sandbox.", "filterable": false, "sortable": true }, { "key": "max_concurrency", "label": "Concurrency", "type": "string", "description": "Published concurrent sandbox limits by plan.", "filterable": false, "sortable": false }, { "key": "soc2", "label": "SOC 2", "type": "enum", "description": "SOC 2 status as stated by the vendor.", "filterable": true, "sortable": false, "enumValues": [ "Type II", "Type I", "stated (type not given)" ] }, { "key": "hipaa", "label": "HIPAA", "type": "bool", "description": "Vendor offers HIPAA-compliant usage/BAA (on any plan).", "filterable": true, "sortable": false }, { "key": "box_cost_hour", "label": "2 vCPU + 4 GiB per hour", "type": "money", "unit": "USD/hr", "description": "Hourly list price of one 2 vCPU / 4 GiB sandbox: 2 x the vCPU rate + 4 x the memory rate, from the two published rates. Left empty for self-hosted software and for vendors without both rates. Vendors that bill CPU only while busy (Vercel, Cloudflare, Upstash, Tensorlake, AWS AgentCore) cost less for idle agents than this shows.", "filterable": true, "sortable": true }, { "key": "vs_e2b", "label": "Price vs E2B", "type": "number", "unit": "x E2B", "description": "The 2 vCPU + 4 GiB hourly price divided by E2B's own ($0.1656 at $0.000014/vCPU-s and $0.0000045/GiB-s). Under 1 is cheaper than E2B.", "filterable": false, "sortable": true }, { "key": "status", "label": "Status", "type": "string", "description": "Preview, experimental, beta, sunset or self-hosting caveats the vendor documents.", "filterable": false, "sortable": false } ], "records": [ { "id": "kubernetes-agent-sandbox", "name": "Agent Sandbox (Kubernetes SIG)", "url": "https://agent-sandbox.sigs.k8s.io", "summary": "Kubernetes-native controller and CRDs for agent sandboxes, delegating isolation to gVisor or Kata Containers, with warm pools for fast allocation. For teams that run their own Kubernetes clusters.", "tags": [ "gvisor", "kata", "kubernetes", "open-source", "self-hosted" ], "fields": { "isolation": "pluggable (gVisor/Kata)", "isolation_options": null, "deployment": "self-hosted", "billing_basis": "free (OSS)", "price_vcpu_hour": null, "price_gib_hour": null, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": true, "free_tier_detail": "Free open-source software (Apache-2.0).", "cold_start_ms": null, "resume_ms": null, "startup_claim": "SandboxWarmPool keeps pre-warmed sandboxes ready to claim (no figure published)", "max_session": null, "persistence": true, "pause_resume": true, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": true, "byoc": null, "open_source": true, "license": "Apache-2.0", "sdks": "Python, Go", "regions": null, "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": null, "hipaa": null, "status": null, "box_cost_hour": null, "vs_e2b": null }, "pricing": { "model": "open-source", "from": 0, "currency": "USD", "note": "Free; you pay for your cluster. Checked 2026-11-09." }, "sources": [ { "url": "https://github.com/kubernetes-sigs/agent-sandbox", "title": "kubernetes-sigs/agent-sandbox (GitHub)", "accessed": "2026-11-09" } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "gke-agent-sandbox", "name": "Agent Sandbox on GKE (Google Cloud)", "url": "https://docs.cloud.google.com/kubernetes-engine/docs/how-to/agent-sandbox", "summary": "Google Kubernetes Engine's managed path for the open-source Agent Sandbox, running AI-generated code in gVisor-isolated pods with warm pools and a Python client. For teams on Google Cloud.", "tags": [ "google-cloud", "gvisor", "hyperscaler", "kubernetes", "managed" ], "fields": { "isolation": "gVisor", "isolation_options": null, "deployment": "managed", "billing_basis": null, "price_vcpu_hour": null, "price_gib_hour": null, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": null, "free_tier_detail": null, "cold_start_ms": null, "resume_ms": null, "startup_claim": "Warm pools let new sandboxes be created \"in under a second\"", "max_session": null, "persistence": true, "pause_resume": null, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": null, "byoc": null, "open_source": true, "license": "Apache-2.0", "sdks": "Python (k8s-agent-sandbox)", "regions": null, "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": null, "hipaa": null, "status": null, "box_cost_hour": null, "vs_e2b": null }, "pricing": { "model": "usage", "note": "No extra charge for Agent Sandbox; standard GKE compute pricing applies. Pod snapshots referenced in docs. Checked 2026-11-09." }, "sources": [ { "url": "https://docs.cloud.google.com/kubernetes-engine/docs/how-to/agent-sandbox", "title": "Isolate AI code execution with Agent Sandbox on GKE", "accessed": "2026-11-09" }, { "url": "https://github.com/kubernetes-sigs/agent-sandbox", "title": "kubernetes-sigs/agent-sandbox (GitHub)", "accessed": "2026-11-09" } ], "verifiedAt": "2026-11-09", "featured": false }, { "id": "alibaba-agent-sandbox", "name": "Alibaba Cloud Agent Sandbox", "url": "https://www.alibabacloud.com/help/en/agent-sandbox/", "summary": "Alibaba Cloud runs each agent sandbox in its own VM, works with the E2B SDK, hibernates idle sandboxes without CPU charges and bills per second. For teams that want E2B-style sandboxes at the lowest list rate, or a Kubernetes API.", "tags": [ "alibaba-cloud", "e2b-compatible", "hibernation", "hyperscaler", "managed" ], "fields": { "isolation": "VM", "isolation_options": "Flash Edition (API, E2B SDK) or Container Edition (Kubernetes API)", "deployment": "managed", "billing_basis": "wall clock", "price_vcpu_hour": 0.017856, "price_gib_hour": 0.008928, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": null, "free_tier_detail": null, "cold_start_ms": null, "resume_ms": null, "startup_claim": "\"Millisecond-level startup\"; hibernated sandboxes resume \"within seconds\" (no figure published)", "max_session": null, "persistence": true, "pause_resume": true, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": "E2B SDK compatible; Kubernetes API (Container Edition)", "regions": "Chinese mainland; Hong Kong and regions outside China", "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": null, "hipaa": null, "status": null, "box_cost_hour": 0.0714, "vs_e2b": 0.43 }, "pricing": { "model": "usage", "from": 0.017856, "unit": "vCPU-hour", "currency": "USD", "note": "Default QoS class, International site, Hong Kong and regions outside China: $0.00000496/vCPU-s and $0.00000248/GiB-s (x3,600). The cheaper Economy class lists $0.00000381 and $0.00000191 per second in the same regions. Reference prices; the console bill is authoritative." }, "sources": [ { "url": "https://www.alibabacloud.com/help/en/agent-sandbox/product-overview/billing-overview", "title": "Alibaba Cloud Agent Sandbox billing overview", "accessed": "2026-11-09", "fields": [ "billing_basis", "pause_resume", "price_gib_hour", "price_vcpu_hour", "regions" ] }, { "url": "https://www.alibabacloud.com/help/en/agent-sandbox/product-overview/agent-sandbox-overview", "title": "What is Agent Sandbox", "accessed": "2026-11-09", "fields": [ "isolation", "isolation_options", "startup_claim", "persistence", "sdks" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "aws-agentcore-code-interpreter", "name": "Amazon Bedrock AgentCore Code Interpreter", "url": "https://aws.amazon.com/bedrock/agentcore/", "summary": "AWS’s managed code-execution tool for agents: each session runs in a dedicated microVM with pre-built Python, JavaScript and TypeScript runtimes, S3 access and CloudTrail logging. For teams building agents on AWS.", "tags": [ "aws", "code-interpreter", "hyperscaler", "managed" ], "fields": { "isolation": "microVM", "isolation_options": null, "deployment": "managed", "billing_basis": "active CPU", "price_vcpu_hour": 0.0895, "price_gib_hour": 0.00945, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": true, "free_tier_detail": "General AWS Free Tier: up to $200 credits for new AWS customers (not Code Interpreter-specific).", "cold_start_ms": null, "resume_ms": null, "startup_claim": null, "max_session": "15 min default, extendable to 8 h", "persistence": null, "pause_resume": null, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": "Executes Python, JavaScript, TypeScript", "regions": null, "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": null, "hipaa": null, "status": null, "box_cost_hour": 0.2168, "vs_e2b": 1.31 }, "pricing": { "model": "usage", "from": 0.0895, "unit": "vCPU-hour", "currency": "USD", "note": "$0.0895/vCPU-h on actual CPU consumption and $0.00945/GB-h on peak memory, billed per second from microVM boot to session termination (I/O wait and idle are free when no background process runs). Files up to 100 MB inline, 5 GB via S3. Pricing checked 2026-11-09." }, "sources": [ { "url": "https://aws.amazon.com/bedrock/agentcore/pricing/", "title": "Amazon Bedrock AgentCore Pricing", "accessed": "2026-11-09" }, { "url": "https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/code-interpreter-tool.html", "title": "AgentCore Code Interpreter developer guide", "accessed": "2026-11-09", "fields": [] }, { "url": "https://docs.aws.amazon.com/bedrock-agentcore/latest/devguide/code-interpreter-session-characteristics.html", "title": "AgentCore Code Interpreter session characteristics", "accessed": "2026-11-11", "fields": [ "isolation", "max_session" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "aws-lambda-microvms", "name": "AWS Lambda MicroVMs", "url": "https://aws.amazon.com/lambda/lambda-microvms/", "summary": "Stateful serverless Firecracker machines for user- and AI-generated code. A separate compute primitive from ordinary Lambda functions.", "tags": [ "aws", "firecracker", "hyperscaler", "managed" ], "fields": { "isolation": "Firecracker microVM", "isolation_options": null, "deployment": "managed", "billing_basis": null, "price_vcpu_hour": null, "price_gib_hour": null, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": null, "free_tier_detail": null, "cold_start_ms": null, "resume_ms": null, "startup_claim": "Launches from a pre-initialized snapshot; AWS gives no numeric start-time SLA.", "max_session": null, "persistence": true, "pause_resume": true, "memory_snapshot": true, "fork": null, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": null, "regions": null, "max_vcpu": 4, "max_memory_gib": 8, "max_concurrency": null, "soc2": null, "hipaa": null, "status": null, "box_cost_hour": null, "vs_e2b": null }, "pricing": { "model": "usage", "note": "Per-second baseline compute while running (default 1 vCPU / 2 GB, up to 4x), active consumption above baseline; snapshots, storage and data transfer extra. Suspended state kept up to 8 hours." }, "sources": [ { "url": "https://aws.amazon.com/lambda/lambda-microvms/", "title": "aws.amazon.com/lambda/lambda-microvms", "accessed": "2026-11-11", "fields": [ "isolation", "persistence", "pause_resume", "memory_snapshot", "startup_claim" ] }, { "url": "https://aws.amazon.com/lambda/pricing/", "title": "aws.amazon.com/lambda/pricing", "accessed": "2026-11-11", "fields": [ "max_vcpu", "max_memory_gib" ] } ], "verifiedAt": "2026-11-11" }, { "id": "azure-dynamic-sessions", "name": "Azure Container Apps Dynamic Sessions", "url": "https://learn.microsoft.com/en-us/azure/container-apps/sessions", "summary": "Azure's Hyper-V-isolated sessions for running LLM-generated code, from prewarmed pools, as a built-in code interpreter or your own container. For teams building agents on Azure.", "tags": [ "azure", "code-interpreter", "hyper-v", "hyperscaler", "managed" ], "fields": { "isolation": "Hyper-V", "isolation_options": "Code interpreter pools or custom container pools", "deployment": "managed", "billing_basis": "session-hour", "price_vcpu_hour": null, "price_gib_hour": null, "price_sandbox_hour_from": 0.03, "paid_plan_from_usd_month": null, "free_tier": null, "free_tier_detail": null, "cold_start_ms": null, "resume_ms": null, "startup_claim": "Prewarmed pools allocate sessions \"in milliseconds\" (no figure published)", "max_session": "Ends after configurable cooldown with no activity", "persistence": null, "pause_resume": null, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": null, "regions": "38 Azure regions (Americas, Europe, APAC, Middle East & Africa)", "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": null, "hipaa": null, "status": null, "box_cost_hour": null, "vs_e2b": null }, "pricing": { "model": "usage", "from": 0.03, "unit": "session-hour", "currency": "USD", "note": "Dynamic Sessions $0.03 per session-hour (East US list price from the Azure Retail Prices API), billed per allocated session in one-hour increments; custom container pools bill on Dedicated plan instances. Checked 2026-11-09." }, "sources": [ { "url": "https://learn.microsoft.com/en-us/azure/container-apps/sessions", "title": "Dynamic sessions in Azure Container Apps", "accessed": "2026-11-09" }, { "url": "https://learn.microsoft.com/en-us/azure/container-apps/billing", "title": "Billing in Azure Container Apps", "accessed": "2026-11-09" }, { "url": "https://prices.azure.com/api/retail/prices?$filter=serviceName%20eq%20%27Azure%20Container%20Apps%27", "title": "Azure Retail Prices API: Container Apps", "accessed": "2026-11-09", "fields": [ "price_sandbox_hour_from" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "beam", "name": "Beam", "url": "https://www.beam.cloud", "summary": "Beam is a serverless GPU and CPU cloud with a Sandboxes product for untrusted code, built on its open-source beta9 engine. For teams that want sandboxes and GPU inference on one bill.", "tags": [ "gpu", "managed", "open-source", "self-hostable" ], "fields": { "isolation": "container", "isolation_options": "microVM with use_vm=True", "deployment": "both", "billing_basis": "wall clock", "price_vcpu_hour": 0.05, "price_gib_hour": 0.01, "price_sandbox_hour_from": 0.06, "paid_plan_from_usd_month": 20, "free_tier": false, "free_tier_detail": "Developer: $20/month, includes $20 in credits each month.", "cold_start_ms": null, "resume_ms": null, "startup_claim": null, "max_session": null, "persistence": null, "pause_resume": null, "memory_snapshot": null, "fork": null, "gpu_support": true, "self_host": true, "byoc": true, "open_source": true, "license": "AGPL-3.0", "sdks": null, "regions": null, "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": "Type II", "hipaa": true, "status": null, "box_cost_hour": 0.14, "vs_e2b": 0.85 }, "pricing": { "model": "usage", "from": 0.06, "unit": "sandbox-hour (1 vCPU, 1 GiB)", "currency": "USD", "note": "Sandboxes from $0.06/h (1 vCPU/1 GiB), billed per ms while running: $0.05/vCPU-h and $0.01/GiB-h; 2 vCPU/8 GiB = $0.18/h. Developer $20/mo with $20 credits, Team $89/mo + usage. BYOC management fee $0.019/vCPU-h and $0.009/GB-h. Pricing checked 2026-11-09." }, "sources": [ { "url": "https://www.beam.cloud/pricing", "title": "Beam Pricing", "accessed": "2026-11-09", "fields": [ "byoc", "hipaa", "isolation", "isolation_options", "self_host", "soc2" ] }, { "url": "https://github.com/beam-cloud/beta9", "title": "Beam beta9 runtime (GitHub)", "accessed": "2026-11-09", "fields": [ "license", "open_source" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "blaxel", "name": "Blaxel", "url": "https://blaxel.ai", "summary": "Blaxel runs agent sandboxes as microVMs that drop to standby after a few seconds idle and resume in about 25 ms with memory intact. For agents with bursty, interactive workloads.", "tags": [ "fast-resume", "managed", "microvm", "standby" ], "fields": { "isolation": "microVM", "isolation_options": null, "deployment": "managed", "billing_basis": "memory while active", "price_vcpu_hour": null, "price_gib_hour": 0.0414, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": true, "free_tier_detail": "Up to $200 free credits on pay-as-you-go.", "cold_start_ms": null, "resume_ms": 25, "startup_claim": "Resume from standby \"in under 25 milliseconds\"; goes to standby about 15 s after the last connection", "max_session": null, "persistence": true, "pause_resume": true, "memory_snapshot": true, "fork": true, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": "TypeScript, Python, Go", "regions": "US and Europe", "max_vcpu": null, "max_memory_gib": 256, "max_concurrency": null, "soc2": "stated (type not given)", "hipaa": true, "status": null, "box_cost_hour": null, "vs_e2b": null }, "pricing": { "model": "usage", "from": 0.0414, "unit": "GB RAM-hour", "currency": "USD", "note": "$0.0000115 per GB RAM-second while active (CPU scales with memory: 8 GB = 4 cores); standby compute $0; snapshots $0.20/GB-month; HIPAA add-on $250/mo. Pricing checked 2026-11-09." }, "sources": [ { "url": "https://blaxel.ai/pricing", "title": "Blaxel Pricing", "accessed": "2026-11-09", "fields": [ "hipaa", "max_memory_gib" ] }, { "url": "https://docs.blaxel.ai/Sandboxes/Overview", "title": "Blaxel Sandboxes overview", "accessed": "2026-11-09", "fields": [ "fork", "isolation", "memory_snapshot", "resume_ms", "sdks", "startup_claim" ] }, { "url": "https://blaxel.ai/", "title": "Blaxel homepage", "accessed": "2026-11-09", "fields": [ "regions", "soc2" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "anthropic-code-execution", "name": "Claude Code Execution Tool (Anthropic)", "url": "https://platform.claude.com/docs/en/agents-and-tools/tool-use/code-execution-tool", "summary": "Anthropic's server-side sandbox lets Claude run Python and Bash inside the API call: a 1 CPU / 5 GiB container with no internet access that you can reuse for 30 days. For agents built on Claude that need data analysis and file generation without running their own sandbox.", "tags": [ "anthropic", "api", "code-interpreter", "llm-api", "managed" ], "fields": { "isolation": "container", "isolation_options": null, "deployment": "managed", "billing_basis": "wall clock", "price_vcpu_hour": null, "price_gib_hour": null, "price_sandbox_hour_from": 0.05, "paid_plan_from_usd_month": null, "free_tier": true, "free_tier_detail": "1,550 free container-hours per organization per month; free when used with web search or web fetch.", "cold_start_ms": null, "resume_ms": null, "startup_claim": null, "max_session": "Containers expire 30 days after creation; checkpointed after ~5 min idle and restored on reuse", "persistence": true, "pause_resume": true, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": "Claude API SDKs; runs Python 3.11 and Bash", "regions": null, "max_vcpu": 1, "max_memory_gib": 5, "max_concurrency": null, "soc2": null, "hipaa": null, "status": null, "box_cost_hour": null, "vs_e2b": null }, "pricing": { "model": "usage", "from": 0.05, "unit": "container-hour (1 CPU, 5 GiB)", "currency": "USD", "note": "$0.05 per container-hour after 1,550 free hours per organization per month, 5-minute minimum; no charge when the request includes web search or web fetch. No internet access inside the container. Pricing checked 2026-11-09." }, "sources": [ { "url": "https://platform.claude.com/docs/en/agents-and-tools/tool-use/code-execution-tool", "title": "Claude code execution tool docs", "accessed": "2026-11-09" } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "cloudflare-sandbox", "name": "Cloudflare Sandbox SDK / Containers", "url": "https://developers.cloudflare.com/sandbox/", "summary": "Cloudflare's Sandbox SDK starts container instances, each a microVM with its own kernel and network, controlled from a Worker. For teams already building on Cloudflare Workers.", "tags": [ "edge", "hyperscaler", "managed", "microvm", "workers" ], "fields": { "isolation": "Firecracker microVM", "isolation_options": "Dynamic Workers (isolates) for JS/Python/Wasm", "deployment": "managed", "billing_basis": "active CPU", "price_vcpu_hour": 0.072, "price_gib_hour": 0.009, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": 5, "free_tier": false, "free_tier_detail": "Workers Paid ($5/mo) includes 375 vCPU-minutes, 25 GiB-hours memory and 200 GB-hours disk per month.", "cold_start_ms": null, "resume_ms": null, "startup_claim": null, "max_session": null, "persistence": true, "pause_resume": null, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": "TypeScript (@cloudflare/sandbox); any language inside the container", "regions": null, "max_vcpu": 4, "max_memory_gib": 12, "max_concurrency": null, "soc2": "Type II", "hipaa": null, "status": null, "box_cost_hour": 0.18, "vs_e2b": 1.09 }, "pricing": { "model": "usage", "from": 0.072, "unit": "vCPU-hour", "currency": "USD", "note": "$0.000020/vCPU-s on active CPU; $0.0000025/GiB-s memory and $0.00000007/GB-s disk on the provisioned instance, beyond Workers Paid allowances; requires Workers Paid ($5/mo). Largest instance standard-4 (4 vCPU, 12 GiB). Pricing checked 2026-11-09." }, "sources": [ { "url": "https://developers.cloudflare.com/sandbox/", "title": "Sandboxes on Cloudflare", "accessed": "2026-11-09" }, { "url": "https://developers.cloudflare.com/containers/pricing/", "title": "Cloudflare Containers pricing", "accessed": "2026-11-09", "fields": [ "price_vcpu_hour", "price_gib_hour", "price_sandbox_hour_from", "paid_plan_from_usd_month", "billing_basis" ] }, { "url": "https://www.cloudflare.com/trust-hub/", "title": "Cloudflare Trust Hub", "accessed": "2026-11-09", "fields": [ "soc2" ] }, { "url": "https://developers.cloudflare.com/sandbox/concepts/", "title": "developers.cloudflare.com/sandbox/concepts", "accessed": "2026-11-11", "fields": [ "isolation" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "daytona", "name": "Daytona", "url": "https://www.daytona.io", "summary": "Daytona runs sandboxes for AI-generated code with a dedicated kernel, filesystem and network per sandbox, plus Linux VM, Windows and GPU options. Built for agent platforms that need fast creation and long-lived sandboxes.", "tags": [ "byoc", "fast-start", "gpu", "managed", "windows" ], "fields": { "isolation": "container", "isolation_options": "Linux VM (pause/resume, memory snapshots, fork), Windows, macOS and GPU sandboxes", "deployment": "managed", "billing_basis": "wall clock", "price_vcpu_hour": 0.0504, "price_gib_hour": 0.0162, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": true, "free_tier_detail": "$200 free compute on pay-as-you-go; startups up to $50k credits.", "cold_start_ms": 90, "resume_ms": null, "startup_claim": "\"Sub 90ms sandbox creation from code to execution\"", "max_session": null, "persistence": true, "pause_resume": true, "memory_snapshot": true, "fork": true, "gpu_support": true, "self_host": null, "byoc": true, "open_source": null, "license": null, "sdks": "Python, TypeScript, Ruby, Go, Java", "regions": "US East, US West, EU Central (Frankfurt), EU West (London), India; GPU in a shared region", "max_vcpu": 4, "max_memory_gib": 8, "max_concurrency": null, "soc2": "Type II", "hipaa": true, "status": null, "box_cost_hour": 0.1656, "vs_e2b": 1 }, "pricing": { "model": "usage", "from": 0.0504, "unit": "vCPU-hour", "currency": "USD", "note": "$0.0504/vCPU-h, $0.0162/GiB-h, storage $0.000108/GiB-h after 5 GiB free, Windows $0.0858/vCPU-h, GPUs from $0.57/h (RTX 4090, preemptible). Standard sandboxes go up to 4 vCPU / 8 GB; each GPU adds up to 16 vCPU and 192 GB. Pricing checked 2026-11-09." }, "sources": [ { "url": "https://www.daytona.io/pricing", "title": "Daytona Pricing", "accessed": "2026-11-09" }, { "url": "https://www.daytona.io/docs/en/sandboxes", "title": "Daytona Sandboxes docs", "accessed": "2026-11-09", "fields": [ "fork", "gpu_support", "isolation", "isolation_options", "max_memory_gib", "max_vcpu", "memory_snapshot", "pause_resume" ] }, { "url": "https://github.com/daytonaio/daytona", "title": "Daytona (GitHub, archived)", "accessed": "2026-11-09" }, { "url": "https://www.daytona.io/", "title": "Daytona homepage", "accessed": "2026-11-09", "fields": [ "byoc", "cold_start_ms", "regions", "startup_claim" ] }, { "url": "https://trust.daytona.io/", "title": "Daytona Trust Center", "accessed": "2026-11-09", "fields": [ "hipaa", "soc2" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "deno-sandbox", "name": "Deno Sandbox", "url": "https://deno.com/sandbox", "summary": "API-driven Linux microVMs for untrusted code, from the Deno team. Deno Deploy, its hosting platform, has an announced shutdown window, so treat it as migration-sensitive.", "tags": [ "managed", "microvm", "sunset-announced" ], "fields": { "isolation": "microVM", "isolation_options": null, "deployment": "managed", "billing_basis": null, "price_vcpu_hour": null, "price_gib_hour": null, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": null, "free_tier_detail": null, "cold_start_ms": null, "resume_ms": null, "startup_claim": "Under one second (vendor claim).", "max_session": "Up to 30 minutes per sandbox session", "persistence": null, "pause_resume": null, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": null, "regions": "Amsterdam, Chicago", "max_vcpu": 2, "max_memory_gib": 4, "max_concurrency": null, "soc2": null, "hipaa": null, "status": "2026-10-09 announcement: Deno team joining Cloudflare; Deno Deploy runs for six months before shutdown.", "box_cost_hour": null, "vs_e2b": null }, "sources": [ { "url": "https://docs.deno.com/sandbox/", "title": "docs.deno.com/sandbox", "accessed": "2026-11-11", "fields": [ "isolation", "startup_claim", "max_session", "regions", "max_vcpu", "max_memory_gib" ] }, { "url": "https://deno.com/blog/cloudflare", "title": "deno.com/blog/cloudflare", "accessed": "2026-11-11", "fields": [ "status" ] } ], "verifiedAt": "2026-11-11" }, { "id": "docker-sandboxes", "name": "Docker Sandboxes", "url": "https://docs.docker.com/ai/sandboxes/", "summary": "Docker's sbx CLI runs AI coding agents in isolated microVM sandboxes on your own machine or on Docker-managed cloud compute. For developers who want to let coding agents run unattended without exposing the host.", "tags": [ "coding-agents", "local", "managed", "microvm" ], "fields": { "isolation": "microVM", "isolation_options": null, "deployment": "both", "billing_basis": null, "price_vcpu_hour": null, "price_gib_hour": null, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": true, "free_tier_detail": "sbx CLI and local sandbox compute free, including commercial use; cloud compute pay-as-you-go.", "cold_start_ms": null, "resume_ms": null, "startup_claim": null, "max_session": null, "persistence": true, "pause_resume": null, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": true, "byoc": null, "open_source": null, "license": null, "sdks": null, "regions": null, "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": null, "hipaa": null, "status": null, "box_cost_hour": null, "vs_e2b": null }, "pricing": { "model": "usage", "from": 0, "currency": "USD", "note": "Local free; cloud compute billed separately (rate not on the docs page). Checked 2026-11-09." }, "sources": [ { "url": "https://docs.docker.com/ai/sandboxes/", "title": "Docker Sandboxes docs", "accessed": "2026-11-09" }, { "url": "https://docs.docker.com/ai/sandboxes/architecture/", "title": "Docker Sandboxes architecture", "accessed": "2026-11-09", "fields": [ "isolation", "persistence" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "fly-machines", "name": "Fly Machines", "url": "https://fly.io/machines/", "summary": "Low-level Firecracker VMs with an API and CLI for lifecycle and sizing. A fit for teams willing to build their own sandbox orchestration on top.", "tags": [ "firecracker", "low-level", "managed" ], "fields": { "isolation": "Firecracker microVM", "isolation_options": null, "deployment": "managed", "billing_basis": null, "price_vcpu_hour": null, "price_gib_hour": null, "price_sandbox_hour_from": 0.003, "paid_plan_from_usd_month": null, "free_tier": null, "free_tier_detail": null, "cold_start_ms": null, "resume_ms": null, "startup_claim": "Existing Machines start and stop at subsecond speeds; image provisioning is separate.", "max_session": null, "persistence": true, "pause_resume": null, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": "REST API, flyctl", "regions": null, "max_vcpu": null, "max_memory_gib": 128, "max_concurrency": null, "soc2": null, "hipaa": null, "status": null, "box_cost_hour": null, "vs_e2b": null }, "pricing": { "model": "usage", "from": 0.003, "unit": "sandbox-hour", "currency": "USD", "note": "Displayed shared-cpu-1x / 256 MB preset at $0.0030/hour; prices depend on region; volumes $0.15/GB-month, snapshots $0.08/GB-month; egress extra." }, "sources": [ { "url": "https://fly.io/machines/", "title": "fly.io/machines", "accessed": "2026-11-11", "fields": [ "isolation" ] }, { "url": "https://fly.io/pricing/", "title": "fly.io/pricing", "accessed": "2026-11-11", "fields": [ "max_memory_gib", "persistence", "price_sandbox_hour_from" ] }, { "url": "https://docs.fly.io/machines", "title": "docs.fly.io/machines", "accessed": "2026-11-11", "fields": [ "sdks", "startup_claim" ] } ], "verifiedAt": "2026-11-11" }, { "id": "fly-sprites", "name": "Fly.io Sprites", "url": "https://fly.io/sprites", "summary": "Sprites are hardware-isolated Linux computers for AI agents on Fly.io, with a persistent 100 GB disk, fast checkpoints and no compute charge while asleep. For long-lived agent workspaces.", "tags": [ "checkpoints", "managed", "persistent", "vm" ], "fields": { "isolation": "VM", "isolation_options": null, "deployment": "managed", "billing_basis": null, "price_vcpu_hour": 0.0385, "price_gib_hour": 0.021875, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": 100, "free_tier": true, "free_tier_detail": "$30 trial credit for new orgs.", "cold_start_ms": null, "resume_ms": null, "startup_claim": "Checkpoints take \"about a second\"; wake time not published", "max_session": null, "persistence": true, "pause_resume": true, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": "JavaScript, Go, Python, Elixir", "regions": null, "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": "Type II", "hipaa": true, "status": null, "box_cost_hour": 0.1645, "vs_e2b": 0.99 }, "pricing": { "model": "usage", "from": 0.0385, "unit": "CPU-hour", "currency": "USD", "note": "$0.0385/CPU-h, $0.021875/GB-h while running; warm/cold Sprites not billed for compute; hot storage ~$0.50/GB-month. Hero plan $100/mo bundles 1,200 CPU-h. Pricing checked 2026-11-09." }, "sources": [ { "url": "https://fly.io/sprites", "title": "Fly.io Sprites", "accessed": "2026-11-09", "fields": [ "sdks" ] }, { "url": "https://fly.io/security", "title": "Fly.io Security", "accessed": "2026-11-09", "fields": [ "hipaa", "soc2" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "freestyle", "name": "Freestyle", "url": "https://www.freestyle.sh", "summary": "Freestyle provides full Linux VMs for AI agents with nested virtualization, hibernation with memory intact and forking. For app-builder and coding-agent products that need real VMs.", "tags": [ "forking", "hibernation", "managed", "vm" ], "fields": { "isolation": "VM", "isolation_options": null, "deployment": "managed", "billing_basis": "wall clock", "price_vcpu_hour": 0.04032, "price_gib_hour": 0.0129, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": 50, "free_tier": true, "free_tier_detail": "Free forever: 200 vCPU-hours and 400 GiB-hours per month, 10 concurrent VMs.", "cold_start_ms": 65, "resume_ms": 100, "startup_claim": "About 65 ms from API request to a ready machine (vendor claim); a suspended VM wakes in under 100 ms with memory restored.", "max_session": "No limit stated (\"can run forever\")", "persistence": true, "pause_resume": true, "memory_snapshot": true, "fork": true, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": null, "regions": null, "max_vcpu": 32, "max_memory_gib": 64, "max_concurrency": "10 (Free), 40 (Hobby), 400 (Pro)", "soc2": null, "hipaa": null, "status": null, "box_cost_hour": 0.1322, "vs_e2b": 0.8 }, "pricing": { "model": "usage", "from": 0.04032, "unit": "vCPU-hour", "currency": "USD", "note": "$0.04032/vCPU-h, $0.0129/GiB-h after included allowances; Hobby $50/mo, Pro $500/mo (fees count toward usage). Pricing checked 2026-11-09." }, "sources": [ { "url": "https://www.freestyle.sh/pricing", "title": "Freestyle Pricing", "accessed": "2026-11-09" }, { "url": "https://www.freestyle.sh/docs", "title": "Freestyle Docs", "accessed": "2026-11-09", "fields": [ "cold_start_ms", "startup_claim" ] }, { "url": "https://www.freestyle.sh/products/vms", "title": "freestyle.sh/products/vms", "accessed": "2026-11-11", "fields": [ "cold_start_ms", "startup_claim", "resume_ms" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "google-agent-platform-code-execution", "name": "Google Agent Platform Code Execution", "url": "https://docs.cloud.google.com/gemini-enterprise-agent-platform/scale/sandbox/code-execution-overview", "summary": "Google's managed code-execution sandbox for agents (formerly Vertex AI Agent Engine): sandboxes start and run code in under a second and keep their state for up to 14 days, with VPC Service Controls, CMEK and HIPAA support. For teams building agents on Google Cloud.", "tags": [ "code-interpreter", "google-cloud", "hyperscaler", "managed" ], "fields": { "isolation": "gVisor", "isolation_options": null, "deployment": "managed", "billing_basis": null, "price_vcpu_hour": 0.085, "price_gib_hour": 0.009, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": null, "free_tier_detail": null, "cold_start_ms": null, "resume_ms": null, "startup_claim": "\"Sandboxes can be created and execute code in under a second\"", "max_session": "State kept up to 14 days (configurable TTL)", "persistence": true, "pause_resume": null, "memory_snapshot": null, "fork": false, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": "Runs Python (pre-installed libraries only)", "regions": null, "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": null, "hipaa": true, "status": null, "box_cost_hour": 0.206, "vs_e2b": 1.24 }, "pricing": { "model": "usage", "from": 0.085, "unit": "vCPU-hour", "currency": "USD", "note": "Billed as Agent Compute ($0.085/vCPU-h) and Agent Memory ($0.009/GiB-h) on allocated resources, rounded to the nearest second. Snapshots are not supported for Code Execution sandboxes; no network access. Pricing checked 2026-11-09." }, "sources": [ { "url": "https://cloud.google.com/vertex-ai/pricing", "title": "Google Cloud Agent Platform pricing", "accessed": "2026-11-09", "fields": [ "price_vcpu_hour", "price_gib_hour" ] }, { "url": "https://docs.cloud.google.com/gemini-enterprise-agent-platform/scale/sandbox/code-execution-overview", "title": "Code Execution overview", "accessed": "2026-11-09", "fields": [ "startup_claim", "max_session", "persistence", "fork", "sdks" ] }, { "url": "https://docs.cloud.google.com/agent-builder/agent-engine/code-execution/overview", "title": "Agent Engine Code Execution (enterprise security)", "accessed": "2026-11-09", "fields": [ "hipaa" ] }, { "url": "https://docs.cloud.google.com/gemini-enterprise-agent-platform/scale/sandbox", "title": "docs.cloud.google.com/gemini-enterprise-agent-platform/scale/sandbox", "accessed": "2026-11-11", "fields": [ "isolation" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "google-cloud-run-sandboxes", "name": "Google Cloud Run Sandboxes", "url": "https://docs.cloud.google.com/run/docs/code-execution", "summary": "Runs untrusted code inside an existing Cloud Run workload. A preview option for teams that already host agents on Cloud Run.", "tags": [ "google-cloud", "hyperscaler", "managed", "preview" ], "fields": { "isolation": "process sandbox", "isolation_options": "No parent environment variables, secrets or metadata-server access by default", "deployment": "managed", "billing_basis": null, "price_vcpu_hour": null, "price_gib_hour": null, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": null, "free_tier_detail": null, "cold_start_ms": null, "resume_ms": null, "startup_claim": null, "max_session": null, "persistence": true, "pause_resume": null, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": "In-container sandbox CLI (create, exec, delete)", "regions": null, "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": null, "hipaa": null, "status": "Preview; subject to pre-GA terms.", "box_cost_hour": null, "vs_e2b": null }, "sources": [ { "url": "https://docs.cloud.google.com/run/docs/code-execution", "title": "docs.cloud.google.com/run/docs/code-execution", "accessed": "2026-11-11", "fields": [ "isolation", "isolation_options", "persistence", "sdks", "status" ] } ], "verifiedAt": "2026-11-11" }, { "id": "hopx", "name": "Hopx", "url": "https://hopx.ai", "summary": "Hopx runs untrusted and AI-generated code in microVM sandboxes that start in about 100 ms, billed per second. For agent builders who want E2B-style sandboxes with a large starting credit.", "tags": [ "fast-start", "firecracker", "managed", "microvm" ], "fields": { "isolation": "Firecracker microVM", "isolation_options": null, "deployment": "managed", "billing_basis": "wall clock", "price_vcpu_hour": 0.0504, "price_gib_hour": 0.0162, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": true, "free_tier_detail": "$200 free credits, no credit card.", "cold_start_ms": 100, "resume_ms": null, "startup_claim": "microVMs \"spin up in ~100ms\"", "max_session": "No execution time limit (\"hours, days, or weeks\")", "persistence": true, "pause_resume": null, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": "Python, JavaScript; MCP server", "regions": null, "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": null, "hipaa": null, "status": null, "box_cost_hour": 0.1656, "vs_e2b": 1 }, "pricing": { "model": "usage", "from": 0.0504, "unit": "vCPU-hour", "currency": "USD", "note": "Published per second: $0.000014/vCPU-s, $0.0000045/GiB-s, $0.00000003/GiB-s storage; hourly values are x3,600. Pricing checked 2026-11-09." }, "sources": [ { "url": "https://hopx.ai/pricing", "title": "Hopx Pricing", "accessed": "2026-11-09" }, { "url": "https://hopx.ai/", "title": "Hopx homepage", "accessed": "2026-11-09", "fields": [ "isolation", "max_session", "persistence", "sdks" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "hugging-face-sandboxes", "name": "Hugging Face Sandboxes", "url": "https://huggingface.co/docs/huggingface_hub/guides/sandbox", "summary": "Cloud sandboxes built on Hub Jobs, with dedicated VMs and cheaper shared pools. The docs recommend dedicated sandboxes for mutually untrusted code.", "tags": [ "experimental", "gpu", "managed" ], "fields": { "isolation": "VM", "isolation_options": "Shared pool: uid + Landlock inside one VM, for same-trust workloads only", "deployment": "managed", "billing_basis": null, "price_vcpu_hour": null, "price_gib_hour": null, "price_sandbox_hour_from": 0.01, "paid_plan_from_usd_month": null, "free_tier": null, "free_tier_detail": null, "cold_start_ms": 6000, "resume_ms": null, "startup_claim": "Dedicated sandbox about 6 seconds; warm-pool creation about one network round-trip.", "max_session": "24 h hard maximum; 10 min default idle timeout", "persistence": null, "pause_resume": null, "memory_snapshot": null, "fork": null, "gpu_support": true, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": null, "regions": null, "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": null, "hipaa": null, "status": "Experimental: API and defaults may change.", "box_cost_hour": null, "vs_e2b": null }, "pricing": { "model": "usage", "from": 0.01, "unit": "sandbox-hour", "currency": "USD", "note": "CPU Basic Job: $0.01/hour for 2 vCPU, 16 GB RAM, 50 GB disk; other Job resources and sandbox networking may add charges." }, "sources": [ { "url": "https://huggingface.co/docs/huggingface_hub/guides/sandbox", "title": "huggingface.co/docs/huggingface_hub/guides/sandbox", "accessed": "2026-11-11", "fields": [ "isolation", "isolation_options", "cold_start_ms", "startup_claim", "max_session", "gpu_support", "status" ] }, { "url": "https://huggingface.co/docs/hub/jobs-pricing", "title": "huggingface.co/docs/hub/jobs-pricing", "accessed": "2026-11-11", "fields": [ "price_sandbox_hour_from" ] } ], "verifiedAt": "2026-11-11" }, { "id": "hyperbrowser", "name": "Hyperbrowser Sandboxes", "url": "https://www.hyperbrowser.ai/sandboxes", "summary": "Isolated VMs for coding agents and dev services alongside Hyperbrowser's browser infrastructure. Usage draws from the same credit balance as its other products.", "tags": [ "browser", "fast-start", "managed" ], "fields": { "isolation": "VM", "isolation_options": null, "deployment": "managed", "billing_basis": "wall clock", "price_vcpu_hour": 0.045, "price_gib_hour": 0.015, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": true, "free_tier_detail": "5,000 credits to start; 1,000 credits = $1 of usage.", "cold_start_ms": 50, "resume_ms": null, "startup_claim": "Under 50 ms (vendor claim).", "max_session": null, "persistence": true, "pause_resume": null, "memory_snapshot": true, "fork": null, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": null, "regions": null, "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": null, "hipaa": null, "status": null, "box_cost_hour": 0.15, "vs_e2b": 0.91 }, "pricing": { "model": "usage", "from": 0.045, "unit": "vCPU-hour", "currency": "USD", "note": "$0.045 per vCPU-hour and $0.015 per allocated GiB-hour while the VM runs." }, "sources": [ { "url": "https://www.hyperbrowser.ai/sandboxes", "title": "hyperbrowser.ai/sandboxes", "accessed": "2026-11-11", "fields": [ "cold_start_ms", "isolation", "memory_snapshot", "persistence", "startup_claim" ] }, { "url": "https://www.hyperbrowser.ai/pricing", "title": "hyperbrowser.ai/pricing", "accessed": "2026-11-11", "fields": [ "billing_basis", "free_tier", "free_tier_detail", "price_gib_hour", "price_vcpu_hour" ] } ], "verifiedAt": "2026-11-11" }, { "id": "judge0", "name": "Judge0", "url": "https://judge0.com/", "summary": "Compile-and-run API for AI code execution, education and assessments. Open-source, shared cloud or dedicated cloud. Built for bounded submissions, not persistent agent machines.", "hq": "HR", "tags": [ "both", "compile-and-run", "open-source" ], "fields": { "isolation": "process sandbox", "isolation_options": "Isolate is the underlying submission sandbox", "deployment": "both", "billing_basis": null, "price_vcpu_hour": null, "price_gib_hour": null, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": null, "free_tier_detail": null, "cold_start_ms": null, "resume_ms": null, "startup_claim": null, "max_session": "CE defaults: 2 s CPU time, 5 s wall time (configurable)", "persistence": null, "pause_resume": null, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": true, "byoc": null, "open_source": true, "license": "GPL-3.0", "sdks": "60+ programming languages executed", "regions": null, "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": null, "hipaa": null, "status": null, "box_cost_hour": null, "vs_e2b": null }, "pricing": { "model": "plan", "note": "Shared-cloud Pro displayed at EUR 27/month for 2,000 submissions/day, EUR 0.001 per extra submission." }, "sources": [ { "url": "https://ce.judge0.com/", "title": "ce.judge0.com", "accessed": "2026-11-11", "fields": [ "isolation", "isolation_options", "max_session" ] }, { "url": "https://judge0.com/", "title": "judge0.com", "accessed": "2026-11-11", "fields": [ "deployment", "sdks" ] }, { "url": "https://raw.githubusercontent.com/judge0/judge0/master/README.md", "title": "github.com", "accessed": "2026-11-11", "fields": [ "license", "open_source", "self_host" ] } ], "verifiedAt": "2026-11-11" }, { "id": "koyeb", "name": "Koyeb Sandboxes", "url": "https://www.koyeb.com", "summary": "Koyeb Sandboxes are isolated environments for AI-generated and untrusted code on Koyeb's serverless platform, with scale-to-zero. For teams that want sandboxes next to their Koyeb apps and GPUs.", "tags": [ "managed", "scale-to-zero" ], "fields": { "isolation": "microVM", "isolation_options": null, "deployment": "managed", "billing_basis": null, "price_vcpu_hour": null, "price_gib_hour": null, "price_sandbox_hour_from": 0.0036, "paid_plan_from_usd_month": null, "free_tier": null, "free_tier_detail": null, "cold_start_ms": null, "resume_ms": null, "startup_claim": null, "max_session": "Auto-delete up to 24 h after creation (12 h after scale-to-zero)", "persistence": null, "pause_resume": true, "memory_snapshot": null, "fork": null, "gpu_support": true, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": "Python, JavaScript", "regions": null, "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": "stated (type not given)", "hipaa": null, "status": null, "box_cost_hour": null, "vs_e2b": null }, "pricing": { "note": "Sandboxes (public preview) need a Starter, Pro ($29/mo + compute) or Scale ($299/mo) plan. Koyeb prices compute per bundled instance, e.g. $0.0144/h for 1 vCPU / 1 GB (small) and $0.0288/h for 2 vCPU / 2 GB (medium); the sandbox docs do not say which instance types sandboxes use, so no per-vCPU rate is listed. Checked 2026-11-09." }, "sources": [ { "url": "https://www.koyeb.com/docs/sandboxes", "title": "Koyeb Sandboxes docs", "accessed": "2026-11-09" }, { "url": "https://www.koyeb.com/docs/sandboxes/sandbox-lifecycle", "title": "Koyeb Sandbox lifecycle", "accessed": "2026-11-09" }, { "url": "https://www.koyeb.com/pricing", "title": "Koyeb Pricing", "accessed": "2026-11-09", "fields": [ "soc2" ] }, { "url": "https://www.koyeb.com/docs/reference/instances", "title": "Koyeb instance types and prices", "accessed": "2026-11-09" }, { "url": "https://www.koyeb.com/blog/koyeb-sandboxes-fast-scalable-fully-isolated-environments-for-ai-agents", "title": "koyeb.com/blog/koyeb-sandboxes-fast-scalable-fully-isolated-environments-for-ai-agents", "accessed": "2026-11-11", "fields": [ "isolation", "price_sandbox_hour_from", "gpu_support" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "microsandbox", "name": "microsandbox", "url": "https://github.com/zerocore-ai/microsandbox", "summary": "Open-source, local-first microVM runtime and library for running untrusted workloads on your own machine or servers. For teams that want self-hosted hardware isolation with no vendor.", "tags": [ "local", "microvm", "open-source", "self-hosted" ], "fields": { "isolation": "microVM", "isolation_options": null, "deployment": "self-hosted", "billing_basis": "free (OSS)", "price_vcpu_hour": null, "price_gib_hour": null, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": true, "free_tier_detail": "Free open-source software (Apache-2.0).", "cold_start_ms": 100, "resume_ms": null, "startup_claim": "\"Average boot times under 100 milliseconds\" (guest boot, M1)", "max_session": null, "persistence": true, "pause_resume": null, "memory_snapshot": true, "fork": null, "gpu_support": null, "self_host": true, "byoc": null, "open_source": true, "license": "Apache-2.0", "sdks": "TypeScript, Rust, Python, Go, Ruby", "regions": "Local: macOS (Apple Silicon), Linux (KVM), Windows (WHP)", "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": null, "hipaa": null, "status": null, "box_cost_hour": null, "vs_e2b": null }, "pricing": { "model": "open-source", "from": 0, "currency": "USD", "note": "Free; you pay for your own hardware. Project labels itself beta. Checked 2026-11-09." }, "sources": [ { "url": "https://github.com/zerocore-ai/microsandbox", "title": "microsandbox (GitHub)", "accessed": "2026-11-09", "fields": [ "memory_snapshot", "persistence" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "modal", "name": "Modal Sandboxes", "url": "https://modal.com", "summary": "Modal's Sandboxes API runs untrusted code in gVisor-isolated containers on Modal's serverless cloud, with GPU access. For Python-first teams that already use Modal or need GPUs next to agent code.", "tags": [ "gpu", "gvisor", "managed", "serverless" ], "fields": { "isolation": "gVisor", "isolation_options": "VM runtime with its own Linux kernel (no GPU)", "deployment": "managed", "billing_basis": "wall clock", "price_vcpu_hour": 0.071, "price_gib_hour": 0.024, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": 250, "free_tier": true, "free_tier_detail": "Starter: $0 plan with $30/month free compute credits.", "cold_start_ms": null, "resume_ms": null, "startup_claim": null, "max_session": "5 min default, up to 24 h", "persistence": true, "pause_resume": null, "memory_snapshot": null, "fork": null, "gpu_support": true, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": "Python, JavaScript, Go", "regions": null, "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": "Type II", "hipaa": true, "status": null, "box_cost_hour": 0.238, "vs_e2b": 1.44 }, "pricing": { "model": "usage", "from": 0.071, "unit": "vCPU-hour", "currency": "USD", "note": "Sandboxes billed at non-preemptible rates (3x list): $0.00003942 per physical core-second (1 core = 2 vCPU) and $0.00000667/GiB-s; normalised to $0.071/vCPU-h and $0.024/GiB-h. HIPAA on Enterprise only. Pricing checked 2026-11-09." }, "sources": [ { "url": "https://modal.com/pricing", "title": "Modal Pricing", "accessed": "2026-11-09" }, { "url": "https://modal.com/docs/guide/sandboxes", "title": "Modal Sandboxes guide", "accessed": "2026-11-09", "fields": [ "gpu_support", "isolation", "isolation_options", "max_session" ] }, { "url": "https://modal.com/docs/guide/security", "title": "Modal security and privacy", "accessed": "2026-11-09", "fields": [ "soc2" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "morph-cloud", "name": "Morph Cloud", "url": "https://cloud.morph.so", "summary": "Morph Cloud runs MorphVM instances and devboxes that can be paused, resumed, snapshotted and branched. For agent and RL workloads that explore many paths from one state.", "tags": [ "branching", "managed", "snapshots" ], "fields": { "isolation": "VM", "isolation_options": null, "deployment": "managed", "billing_basis": "bundled units", "price_vcpu_hour": null, "price_gib_hour": null, "price_sandbox_hour_from": 0.05, "paid_plan_from_usd_month": 40, "free_tier": true, "free_tier_detail": "Free tier: $0/month, 0 starting MCUs (compute must be purchased).", "cold_start_ms": null, "resume_ms": null, "startup_claim": null, "max_session": null, "persistence": true, "pause_resume": true, "memory_snapshot": null, "fork": true, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": null, "regions": null, "max_vcpu": 1024, "max_memory_gib": 4096, "max_concurrency": "8 (Free), 32 (Team), 128 (Scale)", "soc2": null, "hipaa": null, "status": null, "box_cost_hour": null, "vs_e2b": null }, "pricing": { "model": "usage", "from": 0.05, "unit": "MCU (1 vCPU-hour + 4 GB RAM-hours + 16 GB disk-hours)", "currency": "USD", "note": "$0.05 per Morph Compute Unit (MCU). One MCU buys 1 vCPU-hour plus 4 GB RAM-hours plus 16 GB disk-hours, so a 1 vCPU / 4 GB box costs $0.05 an hour; listed under sandbox-hour, not split into per-vCPU and per-GiB rates. Team $40/mo, Scale $250/mo with discounted MCUs. Pricing checked 2026-11-09." }, "sources": [ { "url": "https://cloud.morph.so/web/pricing", "title": "Morph Cloud Pricing", "accessed": "2026-11-09", "fields": [ "max_concurrency", "max_memory_gib", "max_vcpu" ] }, { "url": "https://cloud.morph.so/docs/documentation/overview", "title": "Morph Cloud docs overview", "accessed": "2026-11-09", "fields": [ "isolation", "pause_resume" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "northflank", "name": "Northflank Sandboxes", "url": "https://northflank.com/product/sandboxes", "summary": "Northflank runs each sandbox in its own Cloud Hypervisor microVM (gVisor for GPU), on Northflank's cloud or inside your own AWS, GCP, Azure or Oracle account. For teams that want no runtime limit and BYOC.", "tags": [ "byoc", "gpu", "long-running", "managed", "microvm" ], "fields": { "isolation": "Cloud Hypervisor microVM", "isolation_options": "gVisor for GPU sandboxes", "deployment": "managed", "billing_basis": "wall clock", "price_vcpu_hour": 0.01667, "price_gib_hour": 0.00833, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": true, "free_tier_detail": "Free Sandbox plan: up to $20/month of resources on Northflank Cloud, no card.", "cold_start_ms": null, "resume_ms": null, "startup_claim": "\"Cold start < 1s\"; example cached image ready in 0.8 s", "max_session": "No time limit (runs until deleted)", "persistence": true, "pause_resume": true, "memory_snapshot": null, "fork": null, "gpu_support": true, "self_host": true, "byoc": true, "open_source": null, "license": null, "sdks": "REST API, CLI", "regions": null, "max_vcpu": 288, "max_memory_gib": 2048, "max_concurrency": null, "soc2": "Type II", "hipaa": true, "status": null, "box_cost_hour": 0.0667, "vs_e2b": 0.4 }, "pricing": { "model": "usage", "from": 0.01667, "unit": "vCPU-hour", "currency": "USD", "note": "Northflank Cloud $0.01667/vCPU-h, $0.00833/GB-h; BYOC $0.02083 per managed vCPU-h plus your cloud bill; GPUs from $0.80/h (L4). Pricing checked 2026-11-09." }, "sources": [ { "url": "https://northflank.com/pricing", "title": "Northflank Pricing", "accessed": "2026-11-09" }, { "url": "https://northflank.com/product/sandboxes", "title": "Northflank Sandboxes", "accessed": "2026-11-09", "fields": [ "gpu_support", "isolation", "isolation_options", "max_memory_gib", "max_session", "max_vcpu", "pause_resume", "sdks", "self_host", "startup_claim" ] }, { "url": "https://northflank.com/security", "title": "Northflank Security", "accessed": "2026-11-09", "fields": [ "hipaa", "soc2" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "novita-sandbox", "name": "Novita Sandbox", "url": "https://novita.ai/", "summary": "Managed stateful environments for coding agents, browser workflows and data analysis, with templates, snapshots and memory-preserving pause.", "tags": [ "managed", "snapshots" ], "fields": { "isolation": null, "isolation_options": null, "deployment": "managed", "billing_basis": "wall clock", "price_vcpu_hour": 0.03528, "price_gib_hour": 0.01152, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": null, "free_tier_detail": null, "cold_start_ms": null, "resume_ms": 1000, "startup_claim": "A paused sandbox typically resumes in about one second (vendor claim).", "max_session": "3 h (Paid), 1 h (Free), up to 4 h on Enterprise", "persistence": true, "pause_resume": true, "memory_snapshot": true, "fork": null, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": "Python, Node.js, CLI", "regions": "US (Phoenix, Virginia)", "max_vcpu": 8, "max_memory_gib": 8, "max_concurrency": "5 (Free), 100 (Paid), 500 (Enterprise)", "soc2": null, "hipaa": null, "status": null, "box_cost_hour": 0.1166, "vs_e2b": 0.7 }, "pricing": { "model": "usage", "from": 0.03528, "unit": "vCPU-hour", "currency": "USD", "note": "$0.0000098/vCPU-s and $0.0000032/GiB-s (x3,600); 512-8192 MiB in 512 MiB steps; CPU and RAM billing stops while paused." }, "sources": [ { "url": "https://docs.novita.ai/guides/sandbox-pricing", "title": "docs.novita.ai/guides/sandbox-pricing", "accessed": "2026-11-11", "fields": [ "price_vcpu_hour", "price_gib_hour", "max_memory_gib", "billing_basis" ] }, { "url": "https://docs.novita.ai/guides/sandbox-overview", "title": "docs.novita.ai/guides/sandbox-overview", "accessed": "2026-11-11", "fields": [ "memory_snapshot", "pause_resume", "persistence", "resume_ms", "startup_claim", "regions" ] }, { "url": "https://docs.novita.ai/guides/sandbox-quota-limit", "title": "Novita Agent Sandbox quotas and limits", "accessed": "2026-11-11", "fields": [ "max_vcpu", "max_session", "max_concurrency" ] }, { "url": "https://docs.novita.ai/guides/sandbox-sdk-and-cli", "title": "Novita Agent Sandbox SDK and CLI", "accessed": "2026-11-11", "fields": [ "sdks" ] } ], "verifiedAt": "2026-11-11" }, { "id": "openai-containers", "name": "OpenAI Containers (Code Interpreter / Hosted Shell)", "url": "https://developers.openai.com/api/docs/guides/tools-code-interpreter", "summary": "OpenAI's hosted containers give models in the Responses API a sandboxed VM for Python and shell, in 1, 4, 16 or 64 GB sizes, created automatically or through the Containers API. For agents built on OpenAI models.", "tags": [ "api", "code-interpreter", "llm-api", "managed", "openai" ], "fields": { "isolation": "VM", "isolation_options": null, "deployment": "managed", "billing_basis": "wall clock", "price_vcpu_hour": null, "price_gib_hour": null, "price_sandbox_hour_from": 0.09, "paid_plan_from_usd_month": null, "free_tier": null, "free_tier_detail": null, "cold_start_ms": null, "resume_ms": null, "startup_claim": null, "max_session": "Expires after 20 minutes without use", "persistence": false, "pause_resume": false, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": "OpenAI API (Responses + Containers API); runs Python", "regions": null, "max_vcpu": null, "max_memory_gib": 64, "max_concurrency": null, "soc2": null, "hipaa": null, "status": null, "box_cost_hour": null, "vs_e2b": null }, "pricing": { "model": "usage", "from": 0.09, "unit": "container-hour (1 GB)", "currency": "USD", "note": "$0.03 (1 GB), $0.12 (4 GB), $0.48 (16 GB), $1.92 (64 GB) per 20-minute session per container, so $0.09/h for 1 GB; eligible sessions bill by the minute with a 5-minute minimum. Model tokens billed separately. Pricing checked 2026-11-09." }, "sources": [ { "url": "https://developers.openai.com/api/docs/pricing", "title": "OpenAI API Pricing", "accessed": "2026-11-09", "fields": [ "price_sandbox_hour_from", "billing_basis" ] }, { "url": "https://developers.openai.com/api/docs/guides/tools-code-interpreter", "title": "OpenAI Code Interpreter guide", "accessed": "2026-11-09", "fields": [ "isolation", "max_session", "persistence", "pause_resume", "sdks", "max_memory_gib" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "opensandbox", "name": "OpenSandbox", "url": "https://github.com/opensandbox-group/OpenSandbox", "summary": "Self-hosted sandbox platform with lifecycle APIs, execution tools and SDKs. Starts on local Docker and scales to Kubernetes.", "tags": [ "firecracker", "kubernetes", "open-source", "self-hosted" ], "fields": { "isolation": "Firecracker microVM", "isolation_options": "Other runtime integrations depend on configuration", "deployment": "self-hosted", "billing_basis": "free (OSS)", "price_vcpu_hour": null, "price_gib_hour": null, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": null, "free_tier_detail": null, "cold_start_ms": 64, "resume_ms": null, "startup_claim": "Fast Sandbox Firecracker creation: 64 ms P50 serial, 125.4 ms P99 with 10 concurrent creations (project benchmark).", "max_session": null, "persistence": true, "pause_resume": true, "memory_snapshot": true, "fork": null, "gpu_support": null, "self_host": true, "byoc": null, "open_source": true, "license": "Apache-2.0", "sdks": null, "regions": null, "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": null, "hipaa": null, "status": null, "box_cost_hour": null, "vs_e2b": null }, "sources": [ { "url": "https://raw.githubusercontent.com/opensandbox-group/OpenSandbox/main/README.md", "title": "github.com", "accessed": "2026-11-11", "fields": [ "billing_basis", "cold_start_ms", "deployment", "isolation", "isolation_options", "license", "memory_snapshot", "open_source", "pause_resume", "persistence", "self_host", "startup_claim" ] } ], "verifiedAt": "2026-11-11" }, { "id": "orgo", "name": "Orgo", "url": "https://www.orgo.ai/", "summary": "Persistent cloud desktops for computer-use and coding agents: a browser, terminal and GUI controls, with programmatic fleet access.", "hq": "US", "founded": 2025, "tags": [ "computer-use", "desktop", "managed" ], "fields": { "isolation": "VM", "isolation_options": null, "deployment": "managed", "billing_basis": null, "price_vcpu_hour": null, "price_gib_hour": null, "price_sandbox_hour_from": 0.055, "paid_plan_from_usd_month": 29, "free_tier": null, "free_tier_detail": null, "cold_start_ms": null, "resume_ms": null, "startup_claim": "Linux computers boot in seconds (vendor claim).", "max_session": null, "persistence": true, "pause_resume": null, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": "REST API, Python and TypeScript SDKs, CLI, MCP, SSH", "regions": null, "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": null, "hipaa": null, "status": null, "box_cost_hour": null, "vs_e2b": null }, "pricing": { "model": "usage", "from": 0.055, "unit": "sandbox-hour", "currency": "USD", "note": "Medium Linux (2 vCPU, 8 GB RAM, 40 GB disk): $0.055/hour running, $0.0027/hour stopped; Starter plan $29/month." }, "sources": [ { "url": "https://www.orgo.ai/", "title": "orgo.ai", "accessed": "2026-11-11", "fields": [ "isolation", "persistence" ] }, { "url": "https://www.orgo.ai/pricing.md", "title": "orgo.ai/pricing.md", "accessed": "2026-11-11", "fields": [ "price_sandbox_hour_from", "paid_plan_from_usd_month" ] }, { "url": "https://www.orgo.ai/about", "title": "orgo.ai/about", "accessed": "2026-11-11", "fields": [ "sdks", "startup_claim" ] } ], "verifiedAt": "2026-11-11" }, { "id": "piston", "name": "Piston", "url": "https://github.com/engineer-man/piston", "summary": "Open-source compile-and-run engine with an HTTP API and many language packages. For self-hosted execution of bounded submissions rather than persistent agent desktops.", "tags": [ "compile-and-run", "open-source", "self-hosted" ], "fields": { "isolation": "process sandbox", "isolation_options": "Isolate inside Docker; separate Linux namespaces per submission", "deployment": "self-hosted", "billing_basis": "free (OSS)", "price_vcpu_hour": null, "price_gib_hour": null, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": null, "free_tier_detail": null, "cold_start_ms": null, "resume_ms": null, "startup_claim": null, "max_session": "Defaults: 3 s run CPU and wall time, 10 s compile timeout", "persistence": null, "pause_resume": null, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": true, "byoc": null, "open_source": true, "license": "MIT", "sdks": null, "regions": null, "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": null, "hipaa": null, "status": null, "box_cost_hour": null, "vs_e2b": null }, "sources": [ { "url": "https://raw.githubusercontent.com/engineer-man/piston/master/readme.md", "title": "github.com", "accessed": "2026-11-11", "fields": [ "billing_basis", "deployment", "isolation", "isolation_options", "license", "max_session", "open_source", "self_host" ] } ], "verifiedAt": "2026-11-11" }, { "id": "prime-intellect-sandboxes", "name": "Prime Intellect Sandboxes", "url": "https://primeintellect.ai/sandboxes", "summary": "Prime Intellect runs each sandbox in a hardware-isolated microVM, with pause and resume that keeps memory and processes, checkpoints you can fork, and thousands of sandboxes starting at once. Built for RL rollouts, evals and agent fleets.", "tags": [ "forking", "managed", "microvm", "rl" ], "fields": { "isolation": "microVM", "isolation_options": null, "deployment": "managed", "billing_basis": "wall clock", "price_vcpu_hour": 0.02, "price_gib_hour": 0.0125, "price_sandbox_hour_from": 0.0335, "paid_plan_from_usd_month": null, "free_tier": null, "free_tier_detail": null, "cold_start_ms": null, "resume_ms": null, "startup_claim": "\"Thousands start within seconds\" (no per-sandbox figure)", "max_session": "60 min default, no lifetime limit when set", "persistence": true, "pause_resume": true, "memory_snapshot": true, "fork": true, "gpu_support": false, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": "Python, CLI (prime)", "regions": null, "max_vcpu": 16, "max_memory_gib": 64, "max_concurrency": "1,024 active sandboxes (default account limit)", "soc2": null, "hipaa": null, "status": null, "box_cost_hour": 0.09, "vs_e2b": 0.54 }, "pricing": { "model": "usage", "from": 0.02, "unit": "vCPU-hour", "currency": "USD", "note": "Launch rates valid through 22 December 2026: $0.02/vCPU-h, $0.0125/GiB-h, disk $0.0002/GiB-h, billed while running; a paused sandbox pays for disk only. 1 vCPU / 1 GiB / 5 GiB disk = $0.0335/h. Images and checkpoints $0.10/GiB-month. GPUs \"coming soon\". Pricing checked 2026-11-09." }, "sources": [ { "url": "https://primeintellect.ai/sandboxes", "title": "Prime Intellect Sandboxes", "accessed": "2026-11-09", "fields": [ "price_sandbox_hour_from", "isolation", "startup_claim" ] }, { "url": "https://docs.primeintellect.ai/sandboxes/overview", "title": "Prime Intellect Sandboxes docs", "accessed": "2026-11-09", "fields": [ "price_vcpu_hour", "price_gib_hour", "billing_basis", "max_session", "persistence", "pause_resume", "memory_snapshot", "fork", "gpu_support", "sdks", "max_vcpu", "max_memory_gib", "max_concurrency" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "pydantic-monty", "name": "Pydantic Monty", "url": "https://pydantic.dev/docs/monty/", "summary": "A Rust-written Python sandbox for model-generated code, as an embeddable package and a commercial server. For small controlled computations, not full Linux environments.", "tags": [ "both", "open-source", "restricted-runtime" ], "fields": { "isolation": "restricted runtime", "isolation_options": "Host access only through explicitly supplied functions and mounts", "deployment": "both", "billing_basis": null, "price_vcpu_hour": null, "price_gib_hour": null, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": null, "free_tier_detail": null, "cold_start_ms": 1, "resume_ms": null, "startup_claim": "New OSS sandbox under 1 ms from an already-running pool; commercial server about 2 ms (project claim).", "max_session": "Configurable memory, execution time and recursion depth limits", "persistence": null, "pause_resume": null, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": true, "byoc": null, "open_source": true, "license": "MIT (OSS package); commercial server sold separately", "sdks": "Python 3.14, documented subset", "regions": null, "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": null, "hipaa": null, "status": null, "box_cost_hour": null, "vs_e2b": null }, "sources": [ { "url": "https://raw.githubusercontent.com/pydantic/monty/main/README.md", "title": "github.com", "accessed": "2026-11-11", "fields": [ "cold_start_ms", "deployment", "isolation", "isolation_options", "license", "open_source", "sdks", "self_host", "startup_claim" ] }, { "url": "https://pydantic.dev/docs/monty/concepts/resource-limits/", "title": "pydantic.dev/docs/monty/concepts/resource-limits", "accessed": "2026-11-11", "fields": [ "max_session" ] } ], "verifiedAt": "2026-11-11" }, { "id": "runloop", "name": "Runloop", "url": "https://www.runloop.ai", "summary": "Runloop Devboxes are VM-isolated sandboxes for AI coding agents, with blueprints, snapshots and suspend/resume. Aimed at teams building and evaluating coding agents.", "tags": [ "byoc", "coding-agents", "managed", "snapshots" ], "fields": { "isolation": "microVM", "isolation_options": null, "deployment": "managed", "billing_basis": "wall clock", "price_vcpu_hour": 0.108, "price_gib_hour": 0.0252, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": 250, "free_tier": true, "free_tier_detail": "$50 usage credit with full Pro access on sign-up, no card required.", "cold_start_ms": null, "resume_ms": null, "startup_claim": "\"Startup to running your first command takes a few seconds\"", "max_session": null, "persistence": true, "pause_resume": true, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": null, "byoc": true, "open_source": null, "license": null, "sdks": "Python, TypeScript", "regions": null, "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": "Type II", "hipaa": true, "status": null, "box_cost_hour": 0.3168, "vs_e2b": 1.91 }, "pricing": { "model": "usage", "from": 0.108, "unit": "vCPU-hour", "currency": "USD", "note": "$0.108/CPU-h, $0.0252/GB-h; suspend/resume on Pro ($250/mo); VPC deploy on Enterprise. Pricing checked 2026-11-09." }, "sources": [ { "url": "https://www.runloop.ai/pricing", "title": "Runloop Pricing", "accessed": "2026-11-09", "fields": [ "hipaa", "soc2" ] }, { "url": "https://docs.runloop.ai/docs/devboxes/overview", "title": "Runloop Devboxes overview", "accessed": "2026-11-09" }, { "url": "https://runloop.ai/pricing", "title": "runloop.ai/pricing", "accessed": "2026-11-11", "fields": [ "isolation" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "sandbox0", "name": "Sandbox0", "url": "https://sandbox0.ai/", "summary": "Cloud computers for agents with a durable, automatically growing writable filesystem. Compute pauses between tasks while files stay.", "tags": [ "gvisor", "managed", "persistent" ], "fields": { "isolation": "gVisor", "isolation_options": null, "deployment": "managed", "billing_basis": null, "price_vcpu_hour": 0, "price_gib_hour": 0.015, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": true, "free_tier_detail": "$5 credit for eligible new users after adding a credit card.", "cold_start_ms": null, "resume_ms": null, "startup_claim": null, "max_session": null, "persistence": true, "pause_resume": true, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": false, "byoc": null, "open_source": null, "license": null, "sdks": null, "regions": null, "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": null, "hipaa": null, "status": "Managed cloud only; self-hosting no longer supported, BYOC planned.", "box_cost_hour": 0.06, "vs_e2b": 0.36 }, "pricing": { "model": "usage", "from": 0.015, "unit": "GiB-hour", "currency": "USD", "note": "$0.015 per configured GiB-hour with no separate CPU charge; retained data $0.02 per GiB-month (shared base images excluded)." }, "sources": [ { "url": "https://sandbox0.ai/", "title": "sandbox0.ai", "accessed": "2026-11-11", "fields": [ "deployment", "isolation", "pause_resume", "persistence", "self_host", "status" ] }, { "url": "https://sandbox0.ai/pricing", "title": "sandbox0.ai/pricing", "accessed": "2026-11-11", "fields": [ "free_tier", "free_tier_detail", "price_gib_hour", "price_vcpu_hour" ] } ], "verifiedAt": "2026-11-11" }, { "id": "tensorlake", "name": "Tensorlake Sandboxes", "url": "https://www.tensorlake.ai", "summary": "Tensorlake runs sandboxes on Firecracker microVMs, billed on active CPU with a 20% floor. Pay-as-you-go credits list $0.07 a core-hour; Pro ($250 a month, 25,000 compute units included) cuts that to $0.042 and memory to $0.009 a GB-hour, $0.120 an hour for a 2 vCPU / 4 GiB box.", "tags": [ "active-cpu-billing", "firecracker", "gpu", "managed" ], "fields": { "isolation": "Firecracker microVM", "isolation_options": null, "deployment": "managed", "billing_basis": "active CPU", "price_vcpu_hour": 0.07, "price_gib_hour": 0.015, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": 250, "free_tier": true, "free_tier_detail": "Free forever: 1 sandbox at a time, 1 vCPU, 1 GB RAM, 2-hour sessions.", "cold_start_ms": null, "resume_ms": null, "startup_claim": null, "max_session": "2 h (Free), 24 h (Usage Credits), unlimited (Pro)", "persistence": true, "pause_resume": true, "memory_snapshot": null, "fork": null, "gpu_support": true, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": null, "regions": null, "max_vcpu": 16, "max_memory_gib": 64, "max_concurrency": "1 (Free), 100 (Usage Credits), 1,000 (Pro)", "soc2": "Type II", "hipaa": true, "status": null, "box_cost_hour": 0.2, "vs_e2b": 1.21 }, "pricing": { "model": "usage", "from": 0.07, "unit": "active core-hour", "currency": "USD", "note": "Usage Credits: $0.07/core-h active CPU (20% minimum), $0.015/GB-h RAM; Pro ($250 incl. 25,000 CU): $0.042 and $0.009. H100 $3.95/GPU-h. Pricing checked 2026-11-09." }, "sources": [ { "url": "https://www.tensorlake.ai/pricing", "title": "Tensorlake Pricing", "accessed": "2026-11-11", "fields": [ "hipaa", "soc2" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "together-code-sandbox", "name": "Together Code Sandbox (CodeSandbox SDK)", "url": "https://www.together.ai/code-sandbox", "summary": "Together AI's code sandboxes for building AI development environments, with hibernation, VM cloning and git-versioned storage. For teams already using Together's inference platform.", "tags": [ "cloning", "hibernation", "managed" ], "fields": { "isolation": "microVM", "isolation_options": null, "deployment": "managed", "billing_basis": null, "price_vcpu_hour": 0.0446, "price_gib_hour": 0.0149, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": null, "free_tier_detail": null, "cold_start_ms": 2700, "resume_ms": 500, "startup_claim": "P95: cold start 2.7 s, snapshot resume 500 ms, VM clone under 1 s", "max_session": null, "persistence": true, "pause_resume": true, "memory_snapshot": true, "fork": true, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": "Python, TypeScript, CLI", "regions": null, "max_vcpu": 16, "max_memory_gib": 128, "max_concurrency": null, "soc2": null, "hipaa": null, "status": null, "box_cost_hour": 0.1488, "vs_e2b": 0.9 }, "pricing": { "model": "usage", "from": 0.0446, "unit": "vCPU-hour", "currency": "USD", "note": "Pricing page lists Code Sandbox at $0.0446 per vCPU and $0.0149 per GiB RAM per hour. Together Code Interpreter is a separate product at $0.03 per 60-minute session. Pricing checked 2026-11-09." }, "sources": [ { "url": "https://www.together.ai/code-sandbox", "title": "Together Code Sandbox", "accessed": "2026-11-09" }, { "url": "https://docs.together.ai/docs/together-code-sandbox", "title": "Together Code Sandbox docs", "accessed": "2026-11-09", "fields": [ "max_vcpu" ] }, { "url": "https://www.together.ai/pricing", "title": "Together AI Pricing", "accessed": "2026-11-09", "fields": [ "price_gib_hour", "price_vcpu_hour" ] }, { "url": "https://github.com/codesandbox/codesandbox-sdk", "title": "github.com/codesandbox/codesandbox-sdk", "accessed": "2026-11-11", "fields": [ "isolation" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "upstash-box", "name": "Upstash Box", "url": "https://upstash.com/docs/box", "summary": "Upstash Box gives agents persistent sandboxes billed on active CPU only, with memory free and boxes that freeze when idle. For agents that idle most of the time.", "tags": [ "active-cpu-billing", "managed", "persistent" ], "fields": { "isolation": "container", "isolation_options": null, "deployment": "managed", "billing_basis": "active CPU", "price_vcpu_hour": 0.1, "price_gib_hour": 0, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": true, "free_tier_detail": "Free: 5 CPU-hours/month, 10 concurrent boxes, 2 vCPU / 4 GB per box, 1-hour idle timeout.", "cold_start_ms": null, "resume_ms": null, "startup_claim": null, "max_session": "Idle timeout 1 h (Free), 6 h (Pay as you go), none (Keep Alive)", "persistence": true, "pause_resume": true, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": "TypeScript (@upstash/box), Python (upstash-box); Node.js and Python runtimes", "regions": null, "max_vcpu": 8, "max_memory_gib": 16, "max_concurrency": "10 (Free), 1,000 soft limit (Pay as you go, Fixed)", "soc2": "Type II", "hipaa": null, "status": null, "box_cost_hour": 0.2, "vs_e2b": 1.21 }, "pricing": { "model": "usage", "from": 0.1, "unit": "active CPU-hour", "currency": "USD", "note": "$0.10 per active CPU-hour (Small); memory free; storage $0.10/GB-month after 10 GB; Keep Alive from $8/box/month. Pricing checked 2026-11-09." }, "sources": [ { "url": "https://upstash.com/docs/box/overall/pricing", "title": "Upstash Box pricing", "accessed": "2026-11-09", "fields": [ "max_concurrency", "max_memory_gib", "max_vcpu", "sdks" ] }, { "url": "https://upstash.com/blog/ai-agent-sandbox-providers-compared-2026", "title": "Upstash: AI Agent Sandboxes Compared (2026)", "accessed": "2026-11-09" }, { "url": "https://upstash.com/docs/common/help/compliance", "title": "Upstash Compliance", "accessed": "2026-11-09", "fields": [ "soc2" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "vercel-sandbox", "name": "Vercel Sandbox", "url": "https://vercel.com/sandbox", "summary": "Vercel Sandbox runs agent-generated code in Firecracker microVMs and bills CPU only while code is actively running. Suited to teams on Vercel and agents that spend most of their time waiting on LLM calls.", "tags": [ "active-cpu-billing", "firecracker", "managed", "persistent" ], "fields": { "isolation": "Firecracker microVM", "isolation_options": null, "deployment": "managed", "billing_basis": "active CPU", "price_vcpu_hour": 0.128, "price_gib_hour": 0.0212, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": null, "free_tier": true, "free_tier_detail": "Hobby: 5 active CPU-hours, 420 GB-hours memory, 5,000 creations per month.", "cold_start_ms": null, "resume_ms": null, "startup_claim": "\"Sandboxes start in milliseconds\" (no figure published)", "max_session": "45 min (Hobby), 24 h (Pro/Enterprise) per session; persistent sandboxes resume indefinitely", "persistence": true, "pause_resume": true, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": "JavaScript/TypeScript, Python, CLI", "regions": "19 regions (default iad1)", "max_vcpu": 32, "max_memory_gib": 64, "max_concurrency": "10 (Hobby), 10,000 (Pro/Enterprise)", "soc2": "Type II", "hipaa": true, "status": null, "box_cost_hour": 0.3408, "vs_e2b": 2.06 }, "pricing": { "model": "usage", "from": 0.128, "unit": "active vCPU-hour", "currency": "USD", "note": "iad1 rates: $0.128 per active CPU-hour, $0.0212 per provisioned GB-hour, $0.60 per 1M creations, snapshots $0.08/GB-month. Pro usage draws on the $20/month plan credit first. Pricing checked 2026-11-09." }, "sources": [ { "url": "https://vercel.com/docs/vercel-sandbox/pricing", "title": "Vercel Sandbox pricing and quotas", "accessed": "2026-11-09" }, { "url": "https://vercel.com/docs/vercel-sandbox", "title": "Vercel Sandbox docs", "accessed": "2026-11-09", "fields": [ "isolation" ] }, { "url": "https://vercel.com/security", "title": "Vercel Security", "accessed": "2026-11-09", "fields": [ "hipaa", "soc2" ] } ], "verifiedAt": "2026-11-11", "featured": false }, { "id": "wasmer", "name": "Wasmer", "url": "https://wasmer.io/", "summary": "Runs WebAssembly/WASIX packages locally and on Wasmer's cloud. A narrower runtime for compatible language packages, not an arbitrary Linux VM replacement.", "tags": [ "both", "runtime", "wasm" ], "fields": { "isolation": "WebAssembly", "isolation_options": null, "deployment": "both", "billing_basis": "active CPU", "price_vcpu_hour": null, "price_gib_hour": null, "price_sandbox_hour_from": null, "paid_plan_from_usd_month": 10, "free_tier": null, "free_tier_detail": null, "cold_start_ms": null, "resume_ms": null, "startup_claim": null, "max_session": null, "persistence": null, "pause_resume": null, "memory_snapshot": null, "fork": null, "gpu_support": null, "self_host": null, "byoc": null, "open_source": null, "license": null, "sdks": null, "regions": null, "max_vcpu": null, "max_memory_gib": null, "max_concurrency": null, "soc2": null, "hipaa": null, "status": null, "box_cost_hour": null, "vs_e2b": null }, "pricing": { "model": "plan", "note": "Edge Pro $10/month; overage $0.10 per active CPU-hour after 6 included hours and $0.01 per provisioned GB-hour after 400 included GB-hours. Cloud hosting pricing, not a Linux sandbox reservation." }, "sources": [ { "url": "https://wasmer.io/", "title": "wasmer.io", "accessed": "2026-11-11", "fields": [ "deployment", "isolation" ] }, { "url": "https://wasmer.io/pricing", "title": "wasmer.io/pricing", "accessed": "2026-11-11", "fields": [ "paid_plan_from_usd_month", "billing_basis" ] } ], "verifiedAt": "2026-11-11" } ] }