How the integration works
Pi reads custom model providers from ~/.pi/agent/models.json. Register the DeepSeek-V4.io base URL and model metadata, then select the provider/model pair inside Pi.
Create and manage it on /developer.
The client talks to its native compatible endpoint.
Provider credentials and routing stay on the server.
Before you start
- Pi coding agent installed
- A paid DeepSeek-V4.io account with an API key from /developer
- Access to ~/.pi/agent/models.json
Configure Pi
{
"providers": {
"deepseek-v4-io": {
"baseUrl": "https://deepseek-v4.io/api/v1",
"apiKey": "$DEEPSEEK_V4_API_KEY",
"api": "openai-completions",
"models": [
{
"id": "deepseek-v4-flash",
"name": "DeepSeek V4 Flash",
"reasoning": true,
"input": ["text"],
"contextWindow": 1048576,
"maxTokens": 65536
}
]
}
}
}export DEEPSEEK_V4_API_KEY="your-deepseek-v4-io-key"
piVerify the setup
The page remains Experimental until these checks pass against the public DeepSeek-V4.io gateway.
- 1Select deepseek-v4-io/deepseek-v4-flash.
- 2Ask Pi to inspect a local file through its tool interface.
- 3Confirm streaming, the tool result, and the final answer complete.
Troubleshooting
Pi cannot read the provider
Validate ~/.pi/agent/models.json and confirm the top-level providers object is present.
API key is missing
Export DEEPSEEK_V4_API_KEY before starting Pi; the $VARIABLE reference is resolved from that environment.
Tool loop stops early
Confirm api is openai-completions and start a new session after changing model metadata.
Sources and support boundary
Configuration fields are checked against the tool publisher's documentation. DeepSeek-V4.io owns this gateway guide and support for its endpoint; it does not represent the tool publisher.
Pi custom models documentationReady to connect Pi?
Create a DeepSeek-V4.io API key, copy the configuration above, and verify one read-only tool call before using it on a real project.
Open Developer Dashboard