Are you an LLM? Read llms.txt for a summary of the docs, or llms-full.txt for the full context.
Skip to content

Providers

Amodal supports multiple LLM providers with a unified runtime interface. In hosted Amodal, model provider credentials, allowed models, and default selections are platform settings rather than amodal.json fields.

Supported Providers

The catalog is served by the platform and may include Anthropic, OpenAI, Google, DeepSeek, Groq, Mistral, xAI, and other configured providers. The exact models, prices, context windows, and credential sources shown in Amodal come from the platform catalog and org-level allowlists.

Configuration

Organization Providers

Open Organization > Model Providers to configure provider access:

  • Enable or disable each provider.
  • Choose a credential source, such as BYOK or Amodal-managed keys when available for the workspace plan.
  • Add API keys or provider base URLs where the provider requires them.
  • Select allowed models and optional default/simple/advanced model IDs.
  • Verify provider credentials before using them in runtime traffic.

Agent Model

Open Agent > Models to inspect the current model, provider verification status, available catalog entries, price/context metadata, and recent usage. Agent-level model selection is operational runtime state surfaced by Amodal; it is not committed into the agent repo.

If no agent-specific model is configured, Amodal uses the platform default model, currently Gemini 2.5 Flash (google / gemini-2.5-flash). Prefer that default for starter agents and smoke tests unless you have a reason to trade latency/cost for a larger model.

Runtime Behavior

The hosted runtime receives the selected model from platform configuration when Amodal resolves a deploy. Local and programmatic runtime usage can still pass model overrides through runtime APIs, but starter agent repos should not include models in amodal.json.

Provider failover and routing policy are platform/runtime concerns. Use the Models page and Cost page to verify which provider/model handled traffic.

Slow responses do not necessarily mean the wrong model is selected. Check the session in Amodal for model, token usage, and tool-call timing. A simple chat turn on Gemini Flash should be quick; a turn that calls a slow external API, retries after a provider Bad Gateway, or runs through a larger configured model will feel much slower.

Multi-Model Comparison

Use Amodal eval runs and experiments to compare providers. Run the same eval suite against each provider and compare quality, latency, and cost before changing production routing.