Deep Agents CLI with AI Gateway

Deep Agents CLI is LangChain's terminal coding agent built on the deepagents framework. You can configure it to use AI Gateway for unified model access and spend monitoring.

Run the Vercel CLI setup command for Deep Agents CLI:

npx vercel ai-gateway setup --agent deepagents

The command writes a vercel-ai-gateway provider into ~/.deepagents/config.toml with api_key_env naming AI_GATEWAY_API_KEY, a /model switcher shortlist in leaderboard spend order, and per-model profile tables carrying display names and token limits. Re-run it to refresh the shortlist.

If you can't use the Vercel CLI, configure Deep Agents CLI manually:

1. Create an API key

   Go to the AI Gateway section in the Vercel dashboard sidebar and click API keys to create a new API key.

2. Export the key

   Export the key in your shell:

   Terminal

   export AI_GATEWAY_API_KEY="your-ai-gateway-api-key"

   To persist the key, add the same line to your shell's startup file.

3. Point Deep Agents at the gateway

   Add a vercel-ai-gateway provider to ~/.deepagents/config.toml:

   ~/.deepagents/config.toml

   [models]
   default = "vercel-ai-gateway:openai/gpt-6-astra"
    
   [models.providers.vercel-ai-gateway]
   display_name = "Vercel AI Gateway"
   api_key_env = "AI_GATEWAY_API_KEY"
   base_url = "https://ai-gateway.vercel.sh/coding-agent/v1"
   class_path = "langchain_openai:ChatOpenAI"
   models = ["openai/gpt-6-astra", "anthropic/claude-sonnet-5"]
    
   [models.providers.vercel-ai-gateway.params]
   use_responses_api = false

   The models array feeds the /model switcher rather than gating requests, so any gateway model ID works. class_path selects LangChain's OpenAI chat client, and use_responses_api = false keeps requests on Chat Completions.

4. Start Deep Agents

   Open a new terminal so the key is loaded, then run:

   Terminal

   dcode

   The /model switcher filters as you type. Use any other gateway model directly with /model vercel-ai-gateway:<gateway-model-id>.

View your usage, spend, and request activity in the AI Gateway section in the Vercel dashboard sidebar. See the observability documentation for more details.

• Browse the models catalog for available models
• Configure spend limits on your API key
• Read the deepagents documentation for subagents and tools