Labs
The organizations behind the models — tracked on the lanes that actually decide outcomes: strategy, people, culture, and money, not just shipped technology.
L = lab-reported ·
A = aggregator · tap a value for all sources
| Entity | Talent flow | Strategy coherence | Trust record | Enterprise focus | Shipping velocity | Safety posture |
|---|---|---|---|---|---|---|
| Anthropic Frontier lab; Claude line; safety-forward positioning; enterprise and API focus. | — | — | — | — | — | — |
| DeepSeek Open-weights frontier challenger; extreme cost efficiency; MIT-licensed releases | — | — | — | — | — | — |
| Google DeepMind Google's consolidated AI unit; Gemini line; research depth plus TPU infrastructu | — | — | — | — | — | — |
| Meta AI Llama line historically open-weights; superintelligence-lab reorg; consumer-prod | — | — | — | — | — | — |
| Mistral European lab; open-weights commitment through 2027; enterprise positioning as th | — | — | — | — | — | — |
| OpenAI Frontier lab; GPT line; consumer scale plus API; non-profit-to-commercial govern | — | — | — | — | — | — |
| Qwen (Alibaba) Alibaba's model family; prolific open-weights releases across sizes; strong loca | — | — | — | — | — | — |
| xAI Grok line; massive fast compute buildout; consumer distribution via X; aggressiv | — | — | — | — | — | — |