Using coding agents means setting up multiple accounts, provisioning API keys, and scattering observability and billing. Now, you can route them through AI Gateway to centralize all of this and add controls, with set up in one command: Any of 200+ models in any agent, including models the agent has no native support for. The gateway routes the request to whichever model you picked, with automatic fallbacks.
One dashboard** for agent traffic**: spend, performance, tokens and more, across the models your agents used.
Budgets**, resets, and expiry on the ** keys agents use to control spend. Team-wide policy your agents can't route around: turn on Zero Data Retention or restrict which providers may serve requests, and it holds for every agent request without editing an agent's config.
Request traces, with the cost, tokens, and model behind each request. To start:
vercel ai-gateway coding-agents setup
connects Claude Code, Codex, OpenCode, and more in one command. It finds the agents on your machine, creates the API key, and writes the gateway URL and your credentials into each agent's own config format.
| Agent | Selecting a model |
|---|---|
| Claude Code, CLI and desktop | |
| Codex, CLI and desktop | |
| OpenCode | |
| Pi | |
| Cline | |
| Cursor | Add gateway slugs under Settings, then Models |
| Hermes | |
| Kilo Code | |
| OpenClaw | Add slugs to the starter list in |
Files are edited in place with their formatting preserved, and setup does not pin to a model. This command supports 9 coding agents to utilize the models in AI Gateway. You can also use the /model
in-line command for a selector with all models: refer to the [models list](https://vercel.com/ai-gateway/models) for the full catalog.
Pass `--yes`
to take the defaults, or set the key up front: