Quickstart

GM is compatible with familiar AI SDKs. For an OpenAI-compatible client, change the base URL and use a GM API key.

Sign up with your email or Google and your account starts with a small credit, so your first requests work with no card on file. Signing in with a Bittensor wallet? Add credit first — see billing.

1. Create an API key

Sign in to the GM dashboard, open API keys, and create a key. Copy it when it is shown and store it in an environment variable:

export GM_API_KEY="gm_live_…"

Do not commit the key or paste it into support messages.

2. Find an available model

The model catalog is public and changes independently of this documentation:

curl -fsSL "https://api.saygm.com/v1/models?api_shape=chat.completions"

Choose an entry whose available field is true, then export its exact id:

export GM_MODEL="gpt-5.5"

The availability value is advisory. The inference request is authoritative if availability changes between discovery and use.

3. Make a request

curl "https://api.saygm.com/v1/chat/completions" \
  -H "Authorization: Bearer $GM_API_KEY" \
  -H "Content-Type: application/json" \
  -d "{
    \"model\": \"$GM_MODEL\",
    \"messages\": [{\"role\": \"user\", \"content\": \"Say gm in five words.\"}]
  }"

The response uses the OpenAI Chat Completions shape. Continue with an SDK or tool integration or read the API overview.

When the starting credit runs out

Requests fail with HTTP 402 once the account has no remaining credit. Add credit by card or crypto from billing.