The best-value AI model per provider and per capability, picked automatically from live pricing and benchmark data.
34 models considered
Claude Sonnet 5 (Adaptive Reasoning, Max Effort)
76 models considered
GPT-5.6 Luna (max)
27 models considered
Gemini 3.5 Flash-Lite
373 models considered
DeepSeek V4 Flash 0731 (Reasoning, Max Effort)
160 models considered
GPT-5.6 Luna (max)
311 models considered
Hy3-preview (Reasoning)