OpenAI-compatible chat — Gommo platform (preview)

Coming to Gommo platform

POST /v1/chat/completions is a self-hosted gateway shim, not Gommo public API. For production today use POST https://api.gommo.net/api/v2/chat — see Chat reference. OpenAI-compatible routing will ship on Gommo.

→ Gommo platform roadmap

Gommo Developer exposes POST /v1/chat/completions — map Gommo chat to the OpenAI Chat Completions API so tools like LangChain, OpenAI SDK, or custom agents can plug in without Gommo-specific payloads.

Auth

Same user token as all Gommo APIs:

Authorization: Bearer <user_access_token>

Get a token via public login:

$domain = $env:GOMMO_API_DOMAIN; if (-not $domain) { throw 'Set GOMMO_API_DOMAIN to the user registration domain (dealer)' }
$body = "[email protected]&password=YOUR_PASSWORD&domain=$domain"
$login = Invoke-RestMethod -Method POST `
  -Uri "https://api.gommo.net/api/apps/go-mmo/auth/login" `
  -ContentType "application/x-www-form-urlencoded" -Body $body
$env:TOKEN = $login.access_token

Or paste a token from /app/token/.

Request

curl -X POST http://localhost:3001/v1/chat/completions \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "gpt-5.5::cheap",
    "messages": [
      { "role": "user", "content": "Hello!" }
    ]
  }'

Field	Notes
messages	Required. system / user / assistant roles supported
model	Optional. Format model::server (Gommo) or omit for GOMMO_CHAT_MODEL default
stream	true → SSE in OpenAI chunk format

system messages are prepended to the user query for upstream compatibility.

Response (non-stream)

Standard OpenAI shape:

{
  "id": "chatcmpl-…",
  "object": "chat.completion",
  "choices": [{
    "message": { "role": "assistant", "content": "…" },
    "finish_reason": "stop"
  }],
  "usage": { "prompt_tokens": 0, "completion_tokens": 0, "total_tokens": 0 }
}

Token usage is not reported by Gommo — fields are 0.

Streaming

curl -N -X POST http://localhost:3001/v1/chat/completions \
  -H "Authorization: Bearer $TOKEN" \
  -H "Content-Type: application/json" \
  -d '{"model":"gpt-5.5::cheap","stream":true,"messages":[{"role":"user","content":"Hi"}]}'

Returns text/event-stream with chat.completion.chunk objects and final data: [DONE].

Models list

curl http://localhost:3001/v1/models \
  -H "Authorization: Bearer $TOKEN"

Returns the configured default chat model from env (GOMMO_CHAT_MODEL).

Self-host shims (operators only)

OpenAI-compatible /v1/chat/completions and JSON chat wrappers exist only when self-hosting this repo — Deploy → Self-host (advanced). Production: use direct Gommo form API above.

Next

→ Chat reference · Gommo public API · OpenAPI