Path B — Full diagnosis
Enter my ethical standards
Move each dial to say how much a dimension matters to you. The DoctorKnow Index recomputes live across every model in the directory.
Start from a profile
Every dimension counted evenly.
Privacy & data handling
×5How much do you care that your prompts are not retained or trained on?
Transparency
×5How much do you care about published weights, data provenance and evals?
Carbon footprint
×5How much do you care about the energy source behind each query?
Water & local impact
×5How much do you care about cooling water draw in drought-exposed regions?
Labor & job displacement
×5How much do you care about data-worker treatment and displacement policy?
Regulatory posture
×5How much do you care that the vendor supports binding AI regulation?
The prescription
OLMo 3
Allen Institute for AI · Fully open data, weights, and training code.
Value alignment
89/100
Only model in the index publishing its full training corpus and compute footprint.
Values swap
The raw performance leader is Claude Sonnet 4.6 (92 perf / 76 values). Swap to Mistral Large 3 and you keep 93% of that performance while lifting your alignment to 80.
Ranked against your standards
Compare all →OLMo 3
Allen Institute for AI
Mistral Large 3
Mistral AI
Claude Sonnet 4.6
Anthropic
Claude Opus 4.6
Anthropic
Gemini 3.6 Flash
Google DeepMind
Gemini 3.1 Pro
Google DeepMind
GPT-5.5 Mini
OpenAI
GPT-5.5
OpenAI
Llama 4.1 405B
Meta
Perplexity
Perplexity AI
Microsoft Copilot
Microsoft
Qwen3 Max
Alibaba Cloud
DeepSeek V4
DeepSeek
Grok 4.2
xAI