LaunchAI ModelsJuly 29, 2026
Moonshot AI releases Kimi K3, tops benchmark leaderboards
Kimi K3 is a 2.8-trillion-parameter open-weights model at $3 per million input tokens, ranking #1 on Frontend Web App Arena and SpreadsheetBench 2 and #2 on the Vals Index, passing GPT 5.6 Sol. Together AI measured it matching Claude Fable 5 on software engineering at ~35% of the price.
15 sources
RT @AfterQuery: Kimi K3 ranks #1 on @AfterQuery's SpreadsheetBench 2, surpassing Claude Fable 5....x.com
On Kimi K3: Its Capabilities And Related Discontentsthezvi.substack.com
Where Kimi K3 puts the balance of power — and bends the trajectory — of the AI ecosystem. www.interc...bsky.app
Kimi K3 is now available on AI Gatewayvercel.com
Moonshot AI Releases Kimi K3: A 2.8 Trillion Parameter Open MoE Model With Kimi Delta Attention and 1M Contextmarktechpost.com
Kimi K3 vs Claude Fable 5: Which Open-Weight Model Wins for Agentic Coding?mindstudio.ai
Anyone tried the Q1 Kimi K3 yet? (555GB)huggingface.co
Kimi K3 ranks #1 on @AfterQuery's SpreadsheetBench 2, surpassing Claude Fable 5reddit.com
Self-hosting Kimi K3: 20% more hardware cost, 20% better task resolutionaistack.imec-int.com
Moonshot AI by email
Get an email when Moonshot AI ships something
More stories today
- Thoughtworks' Kief Morris: humans must stay 'on the loop' in AI delivery
- GEMA wins major copyright ruling against Suno, orders damages paid
- LangChain builds ReviewBench benchmark for code review agents
- DeepSeek Flash 0731's reasoning trace amuses with 'OH MY GOD' outburst
- Former OpenAI VP Jerry Tworek discusses AI lab automation