Back to research
Tool · continuously updated

AI Model Tracker

Each brand’s strongest model, one per row — Artificial Analysis Intelligence Index, supply risk, licensing and local runnability. The list is updated regularly.

Last updated · Jul 12, 2026
Model Provider Type License Supply risk Benchmark Downloads (HF) Local
Claude Fable 5 Anthropic Frontier Proprietary High 60 · ·
GPT-5.6 Sol OpenAI Frontier Proprietary High 59 · ·
Gemini 3.5 Flash Google Frontier Proprietary Med 55 · ·
Grok 4.5 xAI Frontier Proprietary Med 54 · ·
GLM-5.2 Z.ai Open-weight MIT Low 51 ·
Qwen3.7 Max Alibaba Frontier Proprietary Med 46 · ·
DeepSeek V4 DeepSeek Open-weight MIT Med 44 ·
Kimi K2.6 Moonshot AI Open-weight MIT Low 44 ·
MiniMax-M3 MiniMax Open-weight Apache-2.0 Low 44 ·
Llama 4 Maverick Meta Open-weight Community Low 42 ·
Nemotron 3 Ultra NVIDIA Open-weight Community Low 38 ·
Mistral Medium 3.5 Mistral AI Frontier Proprietary Med 30 · ·

Supply risk weighs API dependency, regional access constraints and provider concentration together.

Benchmark: the Artificial Analysis Intelligence Index (0-100). Each brand is represented by its strongest model on the site. Source: artificialanalysis.ai.