Skip to main content

Models

celeris-1​

celeris-1 is a diffusion language model optimized for short, structured responses.

Unlike autoregressive LLMs, which produce output one token at a time, a diffusion model can refine multiple output tokens in parallel. This approach is well suited to classification, extraction, scoring, and query rewriting. Typical short requests complete in approximately 50–150 ms within the serving region. Measure your workload with the Server-Timing header.

Use another model for long-form generation. Celeris is intended for concise outputs in latency-sensitive workflows.

PropertyValue
Model name (model field)celeris-1
Context window4,096 tokens (prompt + completion combined)
Best atClassification, extraction, scoring/judging, query rewriting, short structured output
Response modeComplete JSON responses or OpenAI-compatible SSE streaming
Regionus-east-1 (N. Virginia)

:::note The context window is a hard ceiling Prompt plus max_tokens must fit within 4,096 tokens. Oversized requests are rejected up front with a 400 rather than silently truncated. See Rate limits for workspace-level request limits. :::

Model routing and base URLs​

The model you're calling is part of the URL path, ahead of the standard OpenAI-style /v1/... suffix:

https://inference.cloud.celeris.ai/<model>/v1

For celeris-1:

https://inference.cloud.celeris.ai/celeris-1/v1

Two things must agree:

  1. the model segment of the URL path, and
  2. the model field in the request body.

A request whose path names an unavailable model returns 404 Not Found. If the path is correct but the model body field does not match, the API returns a validation error.

You can list what a base URL serves with the standard models endpoint:

curl https://inference.cloud.celeris.ai/celeris-1/v1/models \
  -H "Authorization: Bearer $CELERIS_API_KEY"

Regions​

Celeris is available in us-east-1 (N. Virginia). The base URL is region-independent. For the lowest network latency, deploy latency-sensitive callers near the available region and use the Server-Timing header to separate server time from network time.