Acta Inference

Use familiar APIs across active models.

Send OpenAI-compatible or Anthropic-compatible requests through eligible suppliers and review the published usage rate for each active model.

Start with the request format you already know.

Change the base URL, select a model from the live catalog, and check which features that model supports.

ContractRequest behaviorBoundary

Model catalog

GET /v1/models lists active slugs, modalities, units, and prices.

The live catalog is authoritative.

OpenAI compatibility

Send model and messages to POST /v1/chat/completions.

Streaming is supported where documented.

Anthropic compatibility

Use Messages and token-counting endpoints with supported tools and streaming.

Bearer or x-api-key authentication.

Images and video

Use image generation or asynchronous video jobs.

Availability remains model-dependent.

Provider failover

Order eligible suppliers by route policy, health, economics, and continuity.

Another eligible supplier can be tried after failure.

BYO endpoints

Add project-scoped compatible endpoints to the route set.

External infrastructure remains externally billed.

Published usage pricing

Language and media models publish their applicable units.

Review current rates through GET /v1/models.

Use the documented request shape.

Choose a slug returned by the live catalog for production use.

Read endpoint details and more samples
Python, OpenAI client
from openai import OpenAI

client = OpenAI(
  base_url="https://multitudes.xyz/v1",
  api_key="acta_sk_…"
)
response = client.chat.completions.create(
  model="meta/llama-3.1-8b",
  messages=[{"role":"user", "content":"Summarize this case"}]
)

Send your first request.

Create an account, verify your email, create an API key, then choose a model from the live catalog.

Get an API key