Repository Public
1{"requested":"auto","model":"claude-opus-5-5","provider":"anthropic","reason":"Opus 5.5 was chosen for a deep request under the balanced strategy: frontier tier matches the frontier target; $4/$20 per MTok; validated with this app; reasoning model; the agent loop is built around this provider.","strategy":"balanced","candidates":[{"model":"claude-opus-5-5","score":1.029,"why":"frontier tier matches the frontier target; $4/$20 per MTok; validated with this app; reasoning model; the agent loop is built around this provider"},{"model":"claude-opus-4-8","score":1.02,"why":"frontier tier matches the frontier target; $5/$25 per MTok; validated with this app; reasoning model; the agent loop is built around this provider"},{"model":"claude-fable-5-2","score":1,"why":"frontier tier matches the frontier target; $8/$40 per MTok; validated with this app; reasoning model; the agent loop is built around this provider"},{"model":"gemini-3.1-pro-preview","score":0.803,"why":"frontier tier matches the frontier target; $2/$12 per MTok; validated with this app; reasoning model"},{"model":"gpt-5.5","score":0.765,"why":"frontier tier matches the frontier target; $5/$30 per MTok; validated with this app; reasoning model"}]}2