E2B
✓ Text source-checked Nov 3, 2026E2B runs the isolated machine for AI agents to execute code and use tools.
Source-checked against the vendor's site · rebuilt from the vendor's own pagesProfile updated Link checked Pricing captured How we rate
In short
E2B provides isolated microVMs for AI agents to execute code and use tools. It supports various LLMs and offers cloud or self-hosted deployment options.
The vendor says that E2B is a cloud infrastructure platform that provides an isolated machine for every AI agent session. The vendor says that each agent session runs in a microVM with an Apache-2.0 runtime and SOC 2 Type II compliance. The sandbox environment includes a shell, filesystem, browser, and network access, booting from a snapshot.
The platform offers APIs for lifecycle management, network configuration, file storage, and observability metrics. Agents can make model calls to Anthropic, OpenAI, or open weights, and call over 200 MCP tools from inside the sandbox.
Deployment options include E2B Cloud in US, EU, and APAC regions, or the user's own AWS, GCP, or Azure account.
According to E2B: The sandbox environment includes a shell, Filesystem, browser, and network access, booting from a snapshot. Its pages list this strength: Includes a full environment with shell, filesystem, browser, and network capabilities.
— The AI Tool Bible editorial team
Pros
- ✅ The vendor says that E2B provides true VM-level isolation with an Apache-2.0 runtime and SOC 2 Type II compliance.
- ✅ Includes a full environment with shell, filesystem, browser, and network capabilities.
- ✅ Supports deployment in E2B Cloud regions or within the user's own AWS, GCP, or Azure accounts.
- ✅ Integrates with multiple model providers and over 200 MCP tools for agent functionality.
Pricing (captured )
| Tier | Monthly | Key features |
|---|---|---|
| Hobby | FREE |
|
| pro | $150/month |
|
| Enterprise | CUSTOM |
|
Prices change; we show the grid as captured on Nov 5, 2026. Check current pricing on the vendor's site →
Facts
- Platforms
- linux
- API available
- Yes (not source-checked)
- Free trial
- Yes (not source-checked)
- Open source
- Yes (not source-checked)
- Model used
- LLM-agnostic (works with GPT-4o, Claude, Mistral, Llama, and self-hosted models) (not source-checked)
- Hacker News
- 27 stories · top 36 points · last Oct 6, 2026 · checked Oct 10, 2026
- Earliest Wayback capture
- 2023-04-04 (earliest archived capture of the site, not a launch date)
- Founded
- 2023 (Wikidata)
- Headquarters
- United States (Wikidata)
Where this tool sits
Use cases
Frequently asked
- How much does E2B cost?
- E2B uses a freemium model. There is a free tier available, and paid plans start from $150 per month. Specific pricing details beyond the starting rate are not documented in the provided information.
- Which AI models work with E2B?
- E2B is LLM-agnostic. It works with GPT-4o, Claude, Mistral, Llama, and self-hosted models. Agents can make model calls to Anthropic, OpenAI, or open weights directly from inside the sandbox environment.
- Can I self-host E2B infrastructure?
- Yes. While E2B Cloud is available in US, EU, and APAC regions, you can also deploy the platform on your own AWS, GCP, or Azure account to maintain control over your infrastructure.
- What security features does E2B offer?
- Each agent session runs in an isolated microVM with an Apache-2.0 runtime. The platform maintains SOC 2 Type II compliance. The sandbox includes a shell, filesystem, browser, and network access, booting from a snapshot.
- What tools can agents use inside the sandbox?
- Agents can call over 200 MCP tools from inside the sandbox. The environment also provides APIs for lifecycle management, network configuration, file storage, and observability metrics to support various agent workflows.