Model catalog

Two open-weight text models: GLM 5.2, a long-context flagship, and Gemma 4 31B, a small, cheap model.

Atlas serves a deliberately short catalog. Every model in it is kept warm, so there is no cold-start tier and no model that is cheap because it is rarely available.

This page is a summary. GET /v1/models is the authoritative catalog — it publishes each model's context length, capabilities, JSON Schema subset, and provisional price as data, and it reflects operator changes without a docs release. Read it from your own code rather than hard-coding the table below.

The models

Model	Weights	Context	Tool calling	Structured output	Price / MTok
`glm-5.2`	GLM 5.2 (NVFP4)	262,144	Yes	Yes	$0.95 in · $3.00 out
`gemma-4-31b`	Gemma 4 31B Instruct (FP8)	8,192	No	Yes	$0.13 in · $0.38 out

All prices are provisional and all beta usage is free. See Pricing.

gemma-4-31b does not support tool calling. A request that sends tools to it is refused with a 400 invalid_request_error carrying the code model_does_not_support_tools — refused before the request is admitted, so a capability mismatch never costs you tokens. Its 8,192-token context covers the prompt and the completion together, and its largest accepted request body is 256 KiB rather than 1 MiB. Check the model's capabilities and context_length rather than assuming the catalog is uniform.

Choosing a model

glm-5.2

The default. Start here unless you know you need something else: it supports tool calling and structured output, and its 256K context fits long documents, multi-step reasoning, and agent loops.

gemma-4-31b

The cheap tier, at about a seventh of the flagship's price. Classification, extraction, routing, short rewrites, and JSON extraction with structured output — high-volume work where unit cost dominates and 8K of context is enough. No tool calling.

The honest advice is to benchmark gemma-4-31b on your own prompts before paying for glm-5.2 on high-volume work. Atlas does not claim a quality ranking that holds across workloads, and the catalog is short precisely so that trying both is cheap.

Reading the catalog

GET /v1/models returns the OpenAI list envelope, with Atlas's published metadata added to each entry. The added keys are additive, so an OpenAI SDK parsing this ignores them.

curl https://api.inference.runatlas.com/v1/models \
  -H "Authorization: Bearer $ATLAS_API_KEY"

{
  "object": "list",
  "data": [
    {
      "id": "glm-5.2",
      "object": "model",
      "created": 1790776932,
      "owned_by": "atlas",
      "context_length": 262144,
      "max_request_bytes": 1048576,
      "capabilities": {
        "tool_calling": true,
        "structured_output": true,
        "json_schema_subset": "atlas-json-schema-subset-v1"
      },
      "pricing": {
        "input_per_mtok": "0.95000000",
        "output_per_mtok": "3.00000000",
        "currency": "USD",
        "provisional": true,
        "reasoning_tokens_billed_as": "output"
      },
      "lifecycle": {
        "status": "active",
        "superseded_by": null,
        "shutdown_at": null
      }
    }
  ]
}

context_lengthinteger

Total tokens across the prompt and the completion.

max_request_bytesinteger

The largest request body accepted for this model. A larger one is refused with 413 and the code request_too_large.

capabilities.tool_callingboolean

Whether the model accepts tools and tool_choice. See Tool calling.

capabilities.structured_outputboolean

Whether the model accepts response_format: json_schema. See Structured outputs.

capabilities.json_schema_subsetstring | null

The identifier of the JSON Schema subset this model's schemas are validated against, or null when the model has no structured-output support. Currently always atlas-json-schema-subset-v1.

pricing.reasoning_tokens_billed_asstring

Always "output". Where a model emits a reasoning prelude, those tokens are counted inside completion_tokens — never add the two figures together.

lifecycleobject

status is active or superseded; superseded_by names the successor identifier; shutdown_at is when a superseded identifier stops serving. See Model lifecycle.

Retrieving one model

curl https://api.inference.runatlas.com/v1/models/glm-5.2 \
  -H "Authorization: Bearer $ATLAS_API_KEY"

An identifier that was never published returns 404 with the code model_not_found. An identifier that has been retired — past its published shutdown date — also returns 404, but with the code model_retired, so you can tell "you typed it wrong" from "this used to work". See Model lifecycle.

Not in the catalog

Atlas serves text-generation models only. There are no embedding, audio, image-generation, or vision models, and no /v1/completions, /v1/embeddings, or fine-tuning endpoints. Requests carrying image content parts are not served by any catalog model.