// PRICING

A little access.
A lot to explore.

An experimental API, open to your next idea.
Start with free access. No account or subscription required.

CURRENT ACCESSNO SUBSCRIPTION
$0

FREE ACCESS

  • All models in the current catalog
  • Text, chat, streaming, and image input
  • OpenAI-compatible API
  • Zero prompt and response retention on this gateway
  • API tokens valid for 30 days
GET AN API TOKEN
SHARED CAPACITYCURRENT LIMITS
10 / min

INFERENCE REQUESTS ACROSS THE SERVICE

  • Rolling 60-second request window
  • Up to 2 concurrent inference requests
  • 3 new API tokens per address per hour
  • Retry after the indicated interval on a 429
READ THE LIMITS
#1 / THE CATALOG

Your choice of model.

Choose from 11 currently listed models. Send the model name with your request, or use the default: minimax-m3.

EXPLORE MODELS ↗
#2 / BEFORE YOU BUILD

A small experiment.

Availability and limits may change. The gateway retains no prompts or responses; only operational metrics and counts. Model providers process requests under their own data policies.

ZERO DATA RETENTION ↗