Netlify
Also known as: Netlify Agent Runners
AI native web development platform where developers and AI coding agents build and deploy together, whose Agent Runners bring Claude Code, Codex, and Gemini into the dashboard with full production context, sandboxed execution, and human approved previews, alongside an AI Gateway, MCP server, and observability.
Netlify, based in San Francisco, is an AI native web development platform where developers and AI agents build and deploy together, and the company that coined the term agent experience. Founded in 2014 by chief executive Matt Biilmann and Christian Bach, it is used by more than nine million developers and teams and trusted by brands including Figma, Mattel, and Riot Games to power their digital experiences. Its thesis is that AI has made writing code fast, but that deploying and running that code has not kept up, and that the advantage now goes to platforms where creation flows directly into production.
Netlify's flagship AI capability, Agent Runners, brings leading coding agents, Claude Code, OpenAI Codex, and Google Gemini, directly into the Netlify dashboard, where they run in isolated sandboxes with full context of a project's codebase, logs, environment variables, build settings, and deploy pipeline. A user describes a change in natural language, the agent works on an automated branch, and the result flows through a preview and approval workflow so nothing touches production until a human approves, with one click rollback. Because the agent knows the environment and not just the code, it can wire up a serverless function, configure auth, or fix a broken build without extra setup, and role based access lets designers, product managers, and marketers safely contribute alongside engineers. Around Agent Runners, the platform adds an AI Gateway with built in credentials for Anthropic, OpenAI, and Google, a Netlify MCP server that brings the platform into tools like Cursor and Claude, observability with plain language build failure explanations and secret scanning, and prerendering that serves fully rendered pages to crawlers and AI agents. A dedicated agent onboarding flow lets autonomous agents build and deploy on Netlify with minimal friction.
Netlify is infrastructure for the agentic web, the place where coding agents run against live projects and ship to production, rather than a single vertical agent, which is how its scoring should be read. Its strengths are rich production context, human in the loop approval and sandboxing, model and agent choice, observability, and extensibility through MCP and a deep developer platform, and it hosts agents rather than driving a browser itself. For a team that wants AI coding agents working safely on live web projects with previews, approvals, and production infrastructure built in, Netlify is a strong and differentiated fit; a team wanting an on premise platform, or a turnkey vertical business agent, will find it an AI native build and deploy layer rather than a finished application.
Vendor details
Canonical URL
https://www.netlify.com
Category
Agent infrastructure
Subcategory
AI native web platform for building and deploying with agents
Funding status
Independent, based in San Francisco, founded in 2014 by chief executive Matt Biilmann and Christian Bach. Netlify is a well funded, venture backed company used by more than nine million developers and teams, and trusted by brands including Figma, Mattel, Riot Games, and Bolt. It coined the term agent experience, launched Agent Runners in October 2025 and prompt to app building through netlify.new in March 2026, and has positioned itself as an AI native web platform where humans and agents build side by side.
Company status
independent
Use cases & customers
Primary use cases
Target customers
Deployment options
Integrations
Netlify brings Claude Code, OpenAI Codex, and Google Gemini into the dashboard through Agent Runners, and its MCP server brings the platform into tools like Cursor and Claude. An AI Gateway supplies built in credentials for Anthropic, OpenAI, and Google, projects deploy from tools like Bolt, Lovable, Windsurf, and Kiro, and it integrates GitHub for continuous deployment along with serverless functions, forms, and auth wired up on demand.
In practice
Your backlog is full of small fixes that never justify pulling a developer off feature work. With Netlify Agent Runners you describe the change, an agent that knows your production environment implements it, and you approve a preview, clearing the backlog in minutes.
AI coding tools write code but stop at prototypes, and validating it against production is painful. Netlify runs agents on your live project in isolated sandboxes, generating a preview on real production infrastructure so there are no surprises at deploy.
You want product managers, designers, and marketers to build and update apps without shadow IT. Netlify's role based access and Internal Builder seats let non engineers run agents inside your existing controls, with every change flowing through review.
Sources & related URLs
Research sources
Research notes
Added via Crunchbase agentic discovery CSV, enriched full fidelity 2026-07-07. Categorized Agent infrastructure as the AI native build and deploy platform, Coding agent secondary for Agent Runners. Infrastructure that hosts and runs coding agents (Claude Code, Codex, Gemini) rather than a single agent; many axes score P from the agent capability lens. Public credit based plans plus enterprise. Independent.
Capability coverage
10.0 / 14 capabilities · 71%
| Integrations & Tool CallingBrings Claude Code, Codex, and Gemini into the dashboard, with a Netlify MCP server, an AI Gateway for Anthropic, OpenAI, and Google, GitHub integration, and serverless functions wired on demand, Netlify docs 2026-07-07 | Full |
|---|---|
| Workflow OrchestrationOrchestrates the build to deploy loop through agent runs, previews, and the deploy pipeline, though multi agent orchestration is not the emphasis, Netlify docs 2026-07-07 | Partial |
| Knowledge Grounding & RAGGives agents full context of a project's codebase, logs, environment variables, build settings, and deploy pipeline, so they know the code and the environment, Netlify docs 2026-07-07 | Full |
| Human Oversight & GuardrailsRuns agents in isolated sandboxes where nothing touches production until a human approves, with preview, approval, and one click rollback plus role based access, Netlify docs 2026-07-07 | Full |
| Security, Identity & GovernanceIsolates agent execution in sandboxes, governs who can run agents with role based access, and scans every deploy for exposed secrets before it ships, Netlify docs 2026-07-07 | Full |
| Observability & AuditabilityProvides observability with request traces, logs, and performance data, plain language build failure explanations, and per run credit and result detail, Netlify docs 2026-07-07 | Full |
| Memory & State PersistencePersists durable project and task state, where a task can span several agent runs and redeploys reuse prior changes, though agent learning memory is not documented, Netlify docs 2026-07-07 | Partial |
| Deployment & Data ResidencyDelivered as a managed cloud platform with a global edge network and enterprise controls, though the platform itself is not self hosted, Netlify docs 2026-07-07 | Partial |
| Prebuilt Agents, Templates & PacksOffers starter prompts, reusable prompt templates, and prepackaged platform primitives like serverless functions and AI Gateway, though not a library of prebuilt agents, Netlify docs 2026-07-07 | Partial |
| Triggers & Channel CoverageInvokes agents on demand from the dashboard, CLI, deploy pages, and mobile, with event based webhook and Slack triggering announced on the roadmap, Netlify docs 2026-07-07 | Partial |
| Model Flexibility & RoutingLets users choose their coding agent among Claude Code, Codex, and Gemini, with an AI Gateway providing credentials for Anthropic, OpenAI, and Google models, Netlify docs 2026-07-07 | Full |
| APIs, SDKs & MCP ExtensibilityExtends through a Netlify MCP server, CLI and APIs, an AI Gateway, deploy integrations with Bolt, Lovable, Windsurf, and Kiro, and a dedicated agent onboarding flow, Netlify docs 2026-07-07 | Full |
| Testing, Debugging & OptimizationValidates agent changes against production like previews before publish, with plain language build failure diagnostics and secret scanning, though not an agent evaluation suite, Netlify docs 2026-07-07 | Partial |
| Browser & Computer UseRuns coding agents in sandboxed execution environments rather than driving a browser or operating a computer interface like a human, Netlify docs 2026-07-07 | Unable to verify |
Pricing
Free tier available; Pro around nineteen dollars per member per month, plus AI credits for agent runs
per member subscription plus AI credits for agent runs and inference
Included quota
Credit based plans include a monthly credit allowance for AI inference; agent runs draw down credits per run.
What is public
Free, Personal, and Pro credit based plans are public, with Pro around nineteen dollars per member per month, while AI credit consumption per agent run and enterprise pricing vary.
Billing mechanics
A per member subscription across Free, Personal, and Pro credit based plans, with AI features like Agent Runners metered in credits for inference on top of the plan fee, and enterprise quoted.
Cost watchouts
Agent runs consume AI inference credits on top of the plan fee, for example an individual build costing around one hundred fifty credits, so heavy AI usage can raise cost well beyond the base subscription.
Variable cost rationale
Agent runs consume AI inference credits per run on top of the base subscription, so cost scales with how much AI development a team does.
Additional watchouts
Model your expected agent run volume, since AI inference credits are consumed per run on top of the per member fee, and confirm enterprise terms if you need higher limits.
Overage / add-ons
AI inference is metered in credits; additional usage beyond the plan allowance draws down or requires more credits.
Sales call required
No — self-serve available
Free / trial
Free credit based plan with AI features on by default
Lowest paid plan
Pro plan around nineteen dollars per member per month, plus AI credits
Key ambiguities
Base plans are public, but total cost depends heavily on AI credit consumption per agent run, which varies by task complexity, and enterprise pricing is quoted.
Related vendors
- AgentOps — Agent observability and reliability platform with broad model and…
- Agno — High-performance agent runtime and framework (formerly Phidata) with…
- Apify — Cloud platform for web scraping and automation with 45,000+ prebuilt…
- Arcade — Authenticated tool calling platform and MCP runtime that handles…
- Arize AI — AI observability and evaluation platform that traces, evaluates, and…
- Beam — Open source serverless sandboxes and GPU cloud for production…