Models / gonka-auto

gonka-auto model API

Use gonka-auto for OpenAI-compatible chat completions on decentralized AI infrastructure with transparent usage-based pricing.

Customer model alias: gonka-auto
Endpoint: POST /v1/chat/completions

What is gonka-auto?

gonka-auto is the stable customer-facing model alias for the active verified decentralized inference route. Applications can keep one model name while gonka.broker manages compatible route changes behind the gateway.

Stable routing

The gateway selects a verified decentralized inference route behind the stable gonka-auto alias. Customers do not need to track provider or model-version changes.

Pricing

gonka-auto is billed at 10 nGONKA per input or output token, equivalent to 0.01 GNK per one million total billed tokens. The published rate can change when the underlying decentralized route changes.

Capabilities

  • OpenAI-compatible chat completions
  • Streaming responses
  • Usage and request identifiers
  • Prepaid balance enforcement and spending limits
  • Gateway filtering of upstream reasoning metadata

Quick request

curl https://api.gonka.broker/v1/chat/completions \
  -H "Authorization: Bearer $GONKA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{"model":"gonka-auto","messages":[{"role":"user","content":"Hello"}]}'

Machine-readable model data

Agents and automated tools can read the model identity, pricing formula, capability list, and verification date from models.json.


Next step: Create an account or read the integration docs.