Skip to main content
Every model in the catalog carries metadata — pricing, context window, max output tokens, capabilities, and modes (which decide whether it shows up as a chat, embeddings, image, or audio model). GoModel assembles it from five sources, strongest first:

​
The sources

  1. Pricing overrides — set per model in the dashboard’s Models page. The top layer for pricing fields only; unset price types keep inheriting. See Cost tracking.
  2. config.yaml metadata — providers.<name>.models entries can attach metadata (pricing, context_window, modes, capabilities, …). Declared fields win field-by-field over everything below; omitted fields inherit. This is the escape hatch for local models: declaring modes: [embedding] also derives the model’s category.
  3. Provider discovery — some providers report metadata in their own model listings, and what a provider says about its own deployment wins over the catalog field by field: Gemini’s supportedGenerationMethods, Cohere’s per-model endpoints and context length, OpenRouter’s architecture modalities and context length, Chutes’ context length, max output, capabilities, and pricing, Ollama’s /api/show capabilities, and llama.cpp’s context window and modalities (see llama.cpp). This matters most for self-hosted servers, where the running process is the only source that can know the real context window. Models declared via configured model lists skip this step.
  4. The model catalog — the ai-model-list registry, fetched from MODEL_LIST_URL (default: the registry’s models.min.json on GitHub) at startup and on every catalog refresh. It supplies the rich defaults — pricing, context windows, capabilities, modes — for most hosted models, matching IDs directly, through aliases, and with release-date suffixes stripped, and it fills in every field the provider above did not report. Wrong or missing data is best fixed by contributing to the registry; use an override for an immediate fix.
  5. ID heuristic — a last-resort name check for models that end up with no modes at all (typical for llama.cpp and LM Studio): IDs containing embed or matching well-known embedding families (bge, e5, gte, minilm) become embedding models, IDs containing rerank become reranking models. Namespaced IDs are matched by their final path segment. When unsure, it claims nothing.

​
What metadata affects

  • Pricing drives cost tracking, budgets, and cost load-balancing. Each priced field remembers its source, so the dashboard can show where a rate came from.
  • Modes and categories drive dashboard grouping only — routing never blocks on them, so /v1/embeddings reaches any model the provider serves.
  • Context window and capabilities are advertised on GET /v1/models (and on GET /v1/models/{model} for a single model) for clients that pick models dynamically.

​
Offline behavior

If the catalog fetch fails or the deployment is air-gapped, the gateway runs normally — only the catalog-supplied defaults (including catalog pricing) are missing. Pricing overrides, config.yaml metadata, provider discovery signals, and the ID heuristic still apply. MODEL_LIST_URL accepts a local file as well as an HTTP URL:
A file is read on startup and on every cache refresh, and re-parsed only when its content changes, so you can update pricing by replacing the file. Because a file involves no network request, it keeps working under GOMODEL_OFFLINE=true, which drops HTTP catalog URLs. Set MODEL_LIST_URL=off to turn the catalog off entirely and declare metadata in config.yaml; see the production guide for the full air-gap checklist.
Last modified on September 12, 2026