Orbyt Labs has released Kimi K3, an open model that scores approximately 57 on the Artificial Analysis Intelligence Index, placing it third behind Fable 5. Kimi K3 also achieved the top position on LMArena's Frontend Code Arena, a significant improvement from its predecessor, K2.6. However, the model still lags behind leading closed models in broad agentic tasks. The full 2.8 trillion parameter weights are expected on July 27, and current benchmarks are largely based on Moonshot-reported data or API access. AI
IMPACT This open-source model release offers a competitive alternative to closed models, potentially impacting pricing and accessibility in the AI market.
RANK_REASON New model release from a notable AI lab. [lever_c_demoted from frontier_release: ic=1 ai=1.0]
- Artificial Analysis Intelligence Index
- Frontend Code Arena
- K2.6
- Kimi K3
- Moonshot
- Orbyt Labs
- US government
AI-generated summary · Google Gemini · from 1 sources. How we write summaries →