E2B
E2B provides secure, Firecracker-microVM-based cloud sandboxes for executing AI-generated code, serving AI agent startups, frontier labs, and Fortune 100 enterprises with a freemium self-serve product, Pro subscriptions, and BYOC enterprise deployments.
- Company typePrivate
- Founded2023
- HeadquartersSan Francisco, United States
- Headcount51–100
- GTM typeB2B
- OfferingSoftware
What E2B does
E2B (legally FoundryLabs, Inc., Delaware) is a cloud infrastructure company that provides secure, isolated sandbox environments purpose-built for AI agents and AI-generated code. Its core product, the E2B Sandbox, runs each workload inside a Firecracker microVM (originally developed by AWS) with sub-200ms cold starts, 24-hour session persistence, and support for Python, JavaScript, Ruby, C++, and other languages. On top of this runtime, E2B exposes a Jupyter-like Code Interpreter SDK and a Desktop Sandbox that gives agents a full Linux virtual computer with Chromium browser, terminal, and filesystem. The platform is LLM-agnostic and ships native integrations with OpenAI, Anthropic, Mistral, Meta Llama, and every major agent framework (LangChain, LlamaIndex, Autogen, OpenAI Agents SDK).
The business operates a tiered model: a free Hobby tier with $100 in usage credits and up to 20 concurrent sandboxes, a Pro tier at $150 per month plus per-second compute usage (vCPU, RAM, storage), and custom-priced Enterprise contracts that include BYOC or on-prem deployment across AWS, GCP, and Azure, RBAC, secrets vault, observability, SLAs, and US/EU data residency for EU AI Act compliance. Go-to-market is hybrid: product-led growth through the self-serve funnel and open-source GitHub distribution, combined with a community/developer-relations motion (12.2K+ GitHub stars, active Discord, $20K credit programs for startups and research) and direct enterprise sales for Fortune 100 deployments. Distribution is amplified through ecosystem channels including OpenAI Agents SDK, Vercel, Cloudflare, Docker MCP, Stripe Projects, and Kernel.
E2B serves hyper-growth AI startups and frontier labs (Perplexity, Hugging Face, Manus, Groq, Genspark, Lindy, LMArena, StackAI, Rogo) alongside a long tail of individual developers, with claimed adoption by 88-94% of Fortune 100 companies. The company is venture-backed ($32M total across an Oct 2024 seed led by Decibel and a Jul 2025 Series A led by Insight Partners), headquartered in San Francisco with engineering roots in Prague, and led by founders Vasek Mlejnsky (CEO) and Tomas (CTO).
E2B firmographics
Firmographics- Name
- E2B
- Legal name
- FoundryLabs, Inc.
- Website
- https://e2b.dev
- Company type
- Private
- Founded year
- 2023
- Operating status
- Operating
- Headcount range
- 51–100 employees
- Short description
- E2B provides secure, Firecracker-microVM-based cloud sandboxes for executing AI-generated code, serving AI agent startups, frontier labs, and Fortune 100 enterprises with a freemium self-serve product, Pro subscriptions, and BYOC enterprise deployments.
- Ownership category
- akta.pro rank
E2B industry classification
Industry- Product category
- Cloud Infrastructure for AI Agents
- NAICS
- Computing Infrastructure Providers, Data Processing, Web Hosting, and Related Services (518210)
- SIC
- Services-Prepackaged Software (7372)
- akta.pro primary industry
- Secure Model Deployment & Runtime Protection (sandboxing, isolation) (HDAAAKAH)
- akta.pro secondary industry
- Hyperscale & OCP Server Designs (HDABADAB)
Keywords
Where E2B is headquartered
LocationHeadquarters
- HQ city
- San Francisco
- HQ country
- United States
- HQ region
- North America
Offices2 records
Markets served
E2B business model
Business model- GTM type
- B2B
- Offering type
- Software
- Cost components
- Technology or R&D, Personnel, Infrastructure, Marketing or Sales, Operations
Revenue model
- Hobby (free tier) usage consumption: Free tier with a one-time $100 in usage credits; users can run up to 20 concurrent sandboxes for up to 1 hour; converts to paid when credits are exhausted.
- Pro subscription + usage: $150/month recurring subscription plus usage costs; includes customizable CPU/RAM, 24-hour sessions, and up to 100 concurrent sandboxes (extendable to 1,100).
- Usage-based compute billing: Per-second billing for vCPU, RAM (GiB/s), and storage while a sandbox is running, e.g., 2 vCPU default at $0.000028/s. Applied on top of the tier.
- Enterprise custom contracts: Custom-priced enterprise plans including BYOC/on-prem, RBAC, Secrets Vault, Sandbox Observability, SLA, and US/EU data residency; "Contact us" sales motion.
- Stripe-based billing distribution: Sandboxes are also distributed/sold through Stripe Projects, enabling embedded billing for partner platforms.
Pricing tiers
| Model | Billing | Price |
|---|---|---|
| Freemium | Pay-as-you-go | Hobby: free tier with $100 usage credits, 1-hour sessions, 20 concurrent sandboxes. |
| Subscription | Monthly | Pro: $150/month + usage-based compute charges, 24-hour sessions, 100 concurrent sandboxes. |
| Usage-based | Pay-as-you-go | Usage: per-second compute costs for vCPU, RAM, and storage. |
| Other | Multi-year contract | Enterprise: custom-priced, BYOC/on-prem, RBAC, Secrets Vault, SLA, US & EU regions. |
Go-to-market motion4 records
Distribution channels8 records
Marketing channels11 records
E2B product offering
Product offeringCore offering
E2B provides an open-source, cloud-based sandbox infrastructure built on Firecracker microVMs that lets AI agents and applications execute untrusted code in secure, isolated Linux environments. Each sandbox spins up in under 200ms, supports any programming language, runs up to 24 hours, and provides real-world tools (browser, terminal, filesystem). Customers consume the platform via SDKs (Python, JS/TS), the Code Interpreter, or the Desktop Sandbox, with options for hosted multi-cloud, BYOC, or on-prem deployment.
Product overview
E2B (operated by FoundryLabs, Inc.) offers a single unified product platform — the E2B Sandbox — built on Firecracker microVMs, exposed through two primary product surfaces: the E2B Code Interpreter (a Jupyter-like SDK for executing AI-generated code in Python and JavaScript) and the E2B Desktop Sandbox (a virtual Linux computer for computer-use agents via the open-source E2B Surf project). On top of this core runtime, E2B offers tiered commercial plans — E2B Hobby (free), E2B Pro ($150/month), and E2B Enterprise (BYOC/on-prem with RBAC, secrets vault, observability, and US/EU data residency for EU AI Act compliance) — as well as ecosystem programs (E2B for Startups, E2B for Research) and integration templates (the Kernel-backed template, E2B MCP integration, and distribution via Stripe Projects). The E2B Cookbook is the public library of production-grade reference implementations pairing the sandbox with major LLM providers (OpenAI, Anthropic, Mistral, Meta Llama, Fireworks, Together AI, IBM WatsonX) and AI frameworks (LangChain, LangGraph, LlamaIndex, Autogen, Next.js). Together, these offerings position E2B as the open-source standard for agentic cloud infrastructure, adopted by 88–94% of the Fortune 100 and powering use cases from deep research and AI data analysis to reinforcement learning and fully autonomous multi-agent systems.
Differentiator
Problem solved
Functional benefit
Products and services
- E2B Sandbox Open-source, secure sandboxed cloud runtime for executing AI-generated code. Each sandbox is a Firecracker microVM that spins up in under 200ms, runs up to 24 hours, supports any language (Python, JavaScript, Ruby, C++), installs packages, and provides tools, terminal, filesystem, and browser access. Used by AI agent developers, AI labs, and enterprises.
- E2B Code Interpreter A Jupyter-like code interpreter experience built on top of the E2B Sandbox, exposed via the @e2b/code-interpreter (JS/TS) and e2b-code-interpreter (Python) SDKs. Supports persistent state, multi-cell execution, and LLM-agnostic tool use for AI agents and LLM-based applications.
- E2B Desktop Sandbox Virtual computer sandbox for computer-use agents, providing a full Linux desktop in the cloud (Chromium browser, file manager, terminal) that LLMs can control via the open-source E2B Surf project. Used for secure computer use and headless browser automation.
- E2B Enterprise Enterprise-grade deployment of the E2B platform with BYOC (AWS, GCP, Azure, on-prem) and self-hosting, role-based access control, secrets vault (planned), sandbox observability (planned), SLA guarantees, and US & EU data residency for EU AI Act compliance. Trusted by 94% of Fortune 100 companies.
Quantifiable outcome
- Customer Manus integrated E2B in half a day and would have needed 3–5 full-time infra engineers and months to build comparable infrastructure in-house.
- +8 more outcomes
Companies that use E2B
Customer profileNamed customers13 records
Segments6 records
Ideal customer profiles4 records
E2B technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- Yes
- API docs
- API detail
Core technology
AI maturity
App detail
Integration19 records
AI capability11 records
Feature10 records
E2B partnerships and signals
Strategic signalPartnerships
Eleven partnerships are on record, tiered core, flagship integration, strategic, core integration and core billing/distribution partner.
- KernelcoreReleased a Kernel-backed template that separates agent runtime from the browser environment to address cold starts, CAPTCHAs, proxy management, and authentication loss in headless browser automation.
- OpenAIflagship integrationE2B is one of the supported sandbox providers in OpenAI's Agents SDK (alongside Cloudflare and Vercel). E2B blog also confirms 'E2B is now supported in the OpenAI Agents SDK' (Apr 15, 2026).
- DockerstrategicDocker and E2B partnered to introduce MCP (Model Context Protocol) support in E2B Sandbox, announced in October 2025.
- Anthropiccore integrationFirst-class Anthropic SDK examples in E2B documentation and Cookbook; supports Claude code interpreter and Claude Firecrawl + E2B integration use cases.
- Mistralcore integrationCookbook examples for AI code execution with Mistral's Codestral in both Python and JavaScript/TS; native SDK examples in E2B docs.
- Metacore integrationLlama 3 / Llama 3.1 code-interpreter integration examples in Cookbook and docs, including CSV analysis and function-calling patterns.
- LangChaincore integrationCookbook examples for LangChain and LangGraph with code interpreting; bundled in E2B SDK examples and tutorials.
- LlamaIndexcore integrationLlamaIndex ReAct agent example with E2B execute_python FunctionTool documented in E2B SDK.
- Vercelcore integrationVercel SDK tab in E2B docs; Vercel Sandbox integrated in OpenAI Agents SDK alongside E2B.
- Cloudflarecore integrationCloudflare is listed as a supported sandbox provider in OpenAI Agents SDK alongside E2B and Vercel.
- Stripecore billing/distribution partnerStripe processes Pro tier payments and is integrated as a distribution channel via Stripe Projects; E2B sandboxes available through Stripe Projects.
Scale indicators13 records
Recent moves7 records
Expansion highlights6 records
E2B competitors and assessment
Company assessmentDirect peers
- Modal Labs: Modal provides serverless cloud compute for AI and data science workloads, including sandboxed code execution and GPU runtime, directly competing with E2B's AI agent sandbox use case among developers and AI startups.
- Daytona: Daytona offers a secure, elastic sandbox runtime (often compared head-to-head with E2B, Modal, and Vercel Sandbox) targeted at AI agent code execution and developer workspaces.
- Replit: Replit provides a cloud IDE and sandboxed code execution environment for developers and AI-driven applications, including agent-style automation, overlapping directly with E2B's code interpreter and developer persona.
Broad incumbents
- Vercel: Vercel (with Vercel Sandbox) is co-listed with E2B inside OpenAI's Agents SDK as a supported sandbox provider, offering cloud code execution tied to its broader frontend/edge platform and targeting the same AI agent developer base.
- Cloudflare: Cloudflare offers sandbox execution via its Workers platform and is co-listed with E2B in OpenAI's Agents SDK, bundling sandbox capability with CDN/edge and DDoS protection to compete for the same agent workloads.
- Docker: Docker partnered with E2B on MCP support but also competes via its own container/Docker Sandbox runtime and Model Context Protocol tooling, serving the same AI agent developer audience.
Emerging players
- Northflank: Northflank offers a container-based cloud platform for deploying and running AI workloads, jobs, and APIs, competing for developer mindshare against E2B's managed sandbox runtime.
- Codesphere: Codesphere provides an all-in-one cloud workspace and IDE with code execution, targeting AI and dev use cases that overlap with E2B's developer sandbox offering.
- Fly.io: Fly.io runs containers close to users and offers a developer cloud platform that can host AI agent code execution, positioning as an alternative infrastructure layer to E2B for AI workloads.
- Render: Render is a unified cloud platform for services, jobs, and containers that competes with E2B for AI-backed application hosting and execution workloads, particularly among SMB and developer-led teams.
Market position
Strengths5 records
Weaknesses5 records
Competitive moat6 records
Key risks6 records
Key highlights7 records
Customer concentration
E2B social profiles
Digital presenceE2B compliance and trust
Trust signalCompliance4 records
E2B financial estimates
Financial estimateRevenue estimate
Valuation estimate
E2B leadership team
Management profileNumber of profiles
Profiles3 records
E2B funding detail
Funding detailFunding overview
Funding rounds3 records
Investors7 records
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
E2B M&A and investment
M&A and investmentM&A
Investments
M&A and investment is available on the Subscription and Enterprise plan.Contact sales →
Frequently asked questions about E2B
What does E2B do?
E2B provides an open-source, cloud-based sandbox infrastructure built on Firecracker microVMs that lets AI agents and applications execute untrusted code in secure, isolated Linux environments. Each sandbox spins up in under 200ms, supports any programming language, runs up to 24 hours, and provides real-world tools (browser, terminal, filesystem). Customers consume the platform via SDKs (Python, JS/TS), the Code Interpreter, or the Desktop Sandbox, with options for hosted multi-cloud, BYOC, or on-prem deployment.
Is E2B a public or private company?
E2B is a private company. It is classified as venture growth investor backed and is currently operating.
When was E2B founded?
E2B was founded in 2023. It employs 51 to 100 people.
Where is E2B based?
E2B is headquartered in San Francisco, United States, in the North America region.
How does E2B make money?
Five revenue lines are on record. Hobby (free tier) usage consumption is the primary driver. The others are pro subscription + usage, usage-based compute billing, enterprise custom contracts and stripe-based billing distribution.
Who are E2B's main competitors?
Direct peers on record are Modal Labs, Daytona and Replit. Broad incumbents are Vercel, Cloudflare and Docker. Emerging players are Northflank, Codesphere, Fly.io and Render.
Does E2B have an API?
Yes. Public developer-facing API enabling programmatic creation and management of secure sandboxed cloud environments (microVMs) for AI code execution. SDKs available in Python and Node.js/TypeScript. Authentication via API key. E2B also offers an MCP server with Docker partnership for Model Context Protocol support, and is integrated as a sandbox provider in the OpenAI Agents SDK (Python GA, TypeScript planned). Documentation at https://e2b.dev/docs. Developer documentation is at e2b.dev/docs.
What industry is E2B in?
E2B's product category is Cloud Infrastructure for AI Agents. Its primary akta.pro industry code is HDAAAKAH, Secure Model Deployment & Runtime Protection (sandboxing, isolation), with a secondary code of HDABADAB, Hyperscale & OCP Server Designs. Its NAICS code is 518210 and its SIC code is 7372.