Auto models route each agent to the best performance-to-price model at this moment, and switch providers when one is down — without picking a model by hand.
Fast / Standard / Complex
Gemini 3.6 Flash
Auto · Standard routed to Gemini 3.6 Flash (best performance-to-price of 31 live models)
Best performance-to-price tradeoff, with live multi-provider failover.
Failover chain: deepseek-free → gpt-4o-mini → codestral → qwen-or
Legacy routing modes
DeepSeek Free
balanced routing for chat
Fallbacks: GPT-4o Mini, Gemini 2.5 Flash, Claude Haiku 4.5, DeepSeek Free