Provider profile
Profiles store credentials and provider-specific settings, then load visible models when ready.
Enterprise
Providers and Models
Providers and models control how agents reason, how fast they respond, and what they cost to run. Admins connect provider profiles, curate visible models, and assign runtime roles for primary reasoning, subagents, summarization, code intelligence, and fallback behavior.
What this page covers
Configure provider profiles for OpenAI, Anthropic, Google, OpenRouter, Azure OpenAI, Bedrock, Vertex, OpenAI-compatible endpoints, and other supported providers.
Assign runtime model roles deliberately: Primary, Subagent, Summarize, CodeIntel, and Fallback.
Test with representative incident prompts and tune for quality, latency, cost, and resilience.
Operational detail
How OpsRabbit supports Providers and Models
Providers and models control how agents reason, how fast they respond, and what they cost to run. Admins connect provider profiles, curate visible models, and assign runtime roles for primary reasoning, subagents, summarization, code intelligence, and fallback behavior.
Role baseline
Use the best quality model for Primary, faster or cheaper models for Subagent/Summarize, and a stable separate provider for Fallback.
Troubleshooting
Needs config, empty model lists, or failed role saves usually point to missing credentials, hidden models, disabled profiles, or provider readiness.
Related pages