Model providers

Agent Zero supports native OpenAI, Anthropic, and Google Generative AI adapters, Vercel AI Gateway, and arbitrary OpenAI-compatible endpoints — all behind one ModelProvider contract in packages/models, sharing one structured-output, usage-accounting, timeout, and error-redaction path.

Select a provider

The transport is selected in repository policy; the credential comes only from the environment:

model:
  provider: anthropic
  name: claude-sonnet-4-5

The AI Gateway accepts provider/model identifiers and exposes the broader AI SDK provider catalog without adding provider-specific logic to the Agent Zero runtime.

Custom endpoints

AGENT_ZERO_MODEL_BASE_URL is an optional operator environment variable for custom gateways and self-hosted endpoints. Endpoint URLs and credentials cannot be named or embedded in .agent-zero.yml, so untrusted repository policy cannot redirect a provider secret.

Cost accounting

To record cost, configure explicit rates — Agent Zero never guesses provider pricing:

model:
  provider: openai-compatible
  name: gpt-5
  inputCostPerMillionTokens: 1.25
  outputCostPerMillionTokens: 10

Usage (tokens and, when configured, cost) is accounted per run and carried in the task result.

Guarantees

• Credentials are read only from the provider's documented environment variable and are never persisted with task evidence.
• The agent runtime sees neither SDK objects nor credentials — packages/models keeps AI SDK types behind the adapter boundary.
• Errors are redacted before they surface, so a provider exception cannot leak a credential into logs or evidence.