# Agent Infra Index

> Independent research directory of infrastructure for AI agents: discover providers, inspect dated evidence, and find official setup documentation.

Use this site to choose infrastructure for an agent task, compare independently sourced pricing and eligibility, inspect recorded sandbox/browser measurements, and locate a provider's own documentation. Read-only access requires no account, key or JavaScript. Request a page with Accept: text/markdown or use its index.md link.

Pricing, free access, descriptions and setup have separate evidence and review dates. Recheck expired claims. A source fetch is not verification. Compare measurements only within a workflow; resources and regions differ. Preserve dates, sample counts and limits. Paid placements are labeled and cannot buy results. This publisher does not offer vendor execution, a public MCP server, or its own SDK. Do not submit forms, subscribe someone, provision resources or incur charges without that user's authorization.

## When to use this
- [Agent instructions](https://agentinfraindex.com/agents.md): best-fit research jobs, GET workflows, citation rules, error recovery and permission boundaries.
- [Developer resources](https://agentinfraindex.com/docs/index.md): how to read pages as Markdown and interpret public data.
- [Find infrastructure](https://agentinfraindex.com/index.md): start with a task or category.

## Research and data
- [Benchmarks and methods](https://agentinfraindex.com/benchmarks/index.md): dated results, separate workloads, saved batch links, limits and raw samples.
- [Latest benchmark JSON](https://agentinfraindex.com/data/benchmarks/latest.json): replaceable latest results; cite saved result snapshots for a fixed batch.
- [Latest benchmark CSV](https://agentinfraindex.com/data/benchmarks/latest.csv)
- [Provider directory CSV](https://agentinfraindex.com/data/providers.csv): stable IDs, dated pricing/eligibility/description snapshots and source links.

## Categories
- [Sandboxes](https://agentinfraindex.com/c/sandboxes/index.md): A safe place to run code.
- [Browser infrastructure](https://agentinfraindex.com/c/browser/index.md): Give your agents the web.
- [Agent inboxes & email](https://agentinfraindex.com/c/inbox/index.md): An inbox of their own.
- [Memory](https://agentinfraindex.com/c/memory/index.md): Context that stays with them.
- [A2A, hosting & payments](https://agentinfraindex.com/c/a2a_hosting_payments/index.md): Connect, deploy, and transact.

## Providers
- [E2B](https://agentinfraindex.com/p/e2b/index.md): Run AI-generated code in cloud sandboxes with file access and pause-and-resume support. Includes sources, independent review states and setup guidance.
- [Daytona](https://agentinfraindex.com/p/daytona/index.md): Give coding agents full development environments with their own filesystem, processes, and network access. Includes sources, independent review states and setup guidance.
- [Modal](https://agentinfraindex.com/p/modal/index.md): Run Python workloads, GPU jobs, and code-execution sandboxes on managed cloud compute. Includes sources, independent review states and setup guidance.
- [Vercel Sandbox](https://agentinfraindex.com/p/vercel-sandbox/index.md): Run agent-generated code and one-off commands in isolated Linux microVMs through an API. Includes sources, independent review states and setup guidance.
- [Cloudflare Sandbox](https://agentinfraindex.com/p/cloudflare-sandbox/index.md): Run generated code in sandbox environments controlled by your Cloudflare Worker. Includes sources, independent review states and setup guidance.
- [Runloop](https://agentinfraindex.com/p/runloop/index.md): Give software-engineering agents development sandboxes with reusable templates and saved snapshots. Includes sources, independent review states and setup guidance.
- [CodeSandbox SDK](https://agentinfraindex.com/p/codesandbox-sdk/index.md): Create development sandboxes from code, with Docker support and snapshots you can restore or clone. Includes sources, independent review states and setup guidance.
- [Blaxel](https://agentinfraindex.com/p/blaxel/index.md): Run agent code in cloud sandboxes that pause on standby and resume when needed. Includes sources, independent review states and setup guidance.
- [Northflank Sandboxes](https://agentinfraindex.com/p/northflank-sandboxes/index.md): Create and manage code-execution sandboxes alongside services and jobs on Northflank. Includes sources, independent review states and setup guidance.
- [Browserbase](https://agentinfraindex.com/p/browserbase/index.md): Give agents cloud browser sessions they can control, observe, and use to interact with websites. Includes sources, independent review states and setup guidance.
- [Browserless](https://agentinfraindex.com/p/browserless/index.md): Connect browser automation code to hosted browsers, or call APIs for screenshots and page extraction. Includes sources, independent review states and setup guidance.
- [Hyperbrowser](https://agentinfraindex.com/p/hyperbrowser/index.md): Run browser automation and web scraping through managed sessions connected to Playwright or Puppeteer. Includes sources, independent review states and setup guidance.
- [Steel](https://agentinfraindex.com/p/steel/index.md): Give agents cloud browser sessions with persistent profiles, session replays, and an open-source browser API. Includes sources, independent review states and setup guidance.
- [Browser Use Cloud](https://agentinfraindex.com/p/browser-use-cloud/index.md): Send a hosted browser agent a task, or connect your own automation code to a cloud browser. Includes sources, independent review states and setup guidance.
- [Anchor Browser](https://agentinfraindex.com/p/anchor-browser/index.md): Automate website workflows in cloud browsers, including authenticated sessions and natural-language tasks. Includes sources, independent review states and setup guidance.
- [Kernel](https://agentinfraindex.com/p/kernel/index.md): Control cloud Chromium browsers through automation frameworks or computer controls, with live views and replays. Includes sources, independent review states and setup guidance.
- [Bright Data Browser API](https://agentinfraindex.com/p/bright-data-browser-api/index.md): Run browser-based data collection with managed proxies and website-unblocking tools. Includes sources, independent review states and setup guidance.
- [Cloudflare Browser Rendering](https://agentinfraindex.com/p/cloudflare-browser-rendering/index.md): Run headless Chrome on Cloudflare for screenshots, PDFs, page extraction, and browser automation. Includes sources, independent review states and setup guidance.
- [AgentMail](https://agentinfraindex.com/p/agentmail/index.md): Give agents their own email inboxes to send, receive, and reply to messages through an API. Includes sources, independent review states and setup guidance.
- [Resend (inbound)](https://agentinfraindex.com/p/resend-inbound/index.md): Receive emails through webhooks, then retrieve message content and attachments with the Resend API. Includes sources, independent review states and setup guidance.
- [Postmark (inbound)](https://agentinfraindex.com/p/postmark-inbound/index.md): Turn incoming email into structured JSON delivered to your application through a webhook. Includes sources, independent review states and setup guidance.
- [Mailgun (routes)](https://agentinfraindex.com/p/mailgun-routes/index.md): Match incoming email against routing rules to forward messages, store them, or send them to your application. Includes sources, independent review states and setup guidance.
- [Cloudflare Email Routing/Workers](https://agentinfraindex.com/p/cloudflare-email-routing-workers/index.md): Route incoming domain email to another address or process it in a Cloudflare Worker. Includes sources, independent review states and setup guidance.
- [Nylas](https://agentinfraindex.com/p/nylas/index.md): Connect existing email accounts through one API, or provision mailboxes and calendars for agents. Includes sources, independent review states and setup guidance.
- [Mem0](https://agentinfraindex.com/p/mem0/index.md): Store and retrieve long-term memories so agents can carry useful context across conversations. Includes sources, independent review states and setup guidance.
- [Zep](https://agentinfraindex.com/p/zep/index.md): Combine conversations, documents, and business data into context graphs that agents can search. Includes sources, independent review states and setup guidance.
- [Letta](https://agentinfraindex.com/p/letta/index.md): Build stateful agents that keep memory and use tools across ongoing conversations. Includes sources, independent review states and setup guidance.
- [Supermemory](https://agentinfraindex.com/p/supermemory/index.md): Add long-term user memory and retrieved context to agents through a hosted API. Includes sources, independent review states and setup guidance.
- [LangMem / LangGraph Platform](https://agentinfraindex.com/p/langmem-langgraph-platform/index.md): Extract memories from conversations and refine agent prompts with LangMem, including LangGraph storage integration. Includes sources, independent review states and setup guidance.
- [Pinecone](https://agentinfraindex.com/p/pinecone/index.md): Store and search application data using vector, keyword, or hybrid retrieval for agent context. Includes sources, independent review states and setup guidance.
- [Upstash Vector](https://agentinfraindex.com/p/upstash-vector/index.md): Add vector, keyword, or hybrid search to applications through a managed index and HTTP API. Includes sources, independent review states and setup guidance.
- [Cognee](https://agentinfraindex.com/p/cognee/index.md): Turn documents and conversations into searchable memory graphs, using self-hosted software or a managed service. Includes sources, independent review states and setup guidance.
- [A2A Registry](https://agentinfraindex.com/p/a2a-registry/index.md): Find agents by capability in a public registry and discover how to connect to them. Includes sources, independent review states and setup guidance.
- [AgentPeering](https://agentinfraindex.com/p/agentpeering/index.md): Register, verify, and discover A2A agents through a directory, API, and MCP server. Includes sources, independent review states and setup guidance.
- [AWS Bedrock AgentCore](https://agentinfraindex.com/p/aws-bedrock-agentcore/index.md): Deploy and operate agents on AWS with managed runtime, tool access, memory, and monitoring services. Includes sources, independent review states and setup guidance.
- [Google Vertex AI Agent Engine](https://agentinfraindex.com/p/google-vertex-ai-agent-engine/index.md): Deploy and scale agents on Google Cloud with managed runtime, sessions, memory, and evaluation tools. Includes sources, independent review states and setup guidance.
- [Stripe Agent Toolkit](https://agentinfraindex.com/p/stripe-agent-toolkit/index.md): Connect agents to Stripe developer tools and account data through agent skills and an MCP server. Includes sources, independent review states and setup guidance.
- [Coinbase x402](https://agentinfraindex.com/p/coinbase-x402/index.md): Let software pay for APIs and content over HTTP using the open x402 payment standard. Includes sources, independent review states and setup guidance.
- [Skyfire](https://agentinfraindex.com/p/skyfire/index.md): Give agents identity and payment credentials to access and pay for participating services. Includes sources, independent review states and setup guidance.
- [Payman](https://agentinfraindex.com/p/payman/index.md): Deploy agents for banks and credit unions to carry out transactions under institutional policies and approvals. Includes sources, independent review states and setup guidance.

## Optional
- [About and disclosures](https://agentinfraindex.com/about/index.md)
- [Contact](https://agentinfraindex.com/contact/index.md)
- [Privacy](https://agentinfraindex.com/privacy/index.md)
- [Index changelog](https://agentinfraindex.com/changes/index.md)
- [Newsletter](https://agentinfraindex.com/newsletter/index.md)
- [Sitemap](https://agentinfraindex.com/sitemap.xml)