Describe what you want to do, mark the constraints that matter, and get a ranked list of the AI models that fit. Covers Anthropic (Claude Opus 4.7 / Sonnet 4.6 / Haiku 4.5 / Claude Code), OpenAI (GPT-5, o4, Codex CLI), Google (Gemini 2.5 Pro / Flash / CLI), Meta Llama 4, Alibaba Qwen 3 (incl. Coder), DeepSeek V3 + R1, Mistral, plus image (Midjourney v7, Ideogram v3, DALL·E 3, Flux 1.1 Pro, SD 3.5, Imagen 3, Adobe Firefly, Recraft v3) and voice (ElevenLabs, Deepgram, Whisper). Already used a model and want to know if it was the right pick? Use the AI Model Fit Audit. Read the deep-dive: Picking the right AI model in 2026.
Step 1 — describe the task
Step 2 — pick the task category
Step 3 — mark constraints (optional)
Each checked constraint either rewards models that satisfy it (+5) or eliminates models that fail it (-10).