SignalTech/AI
Kimi K3.1 Pro
High-priority alert based on source significance and keyword relevance.
Context
What happened: Moonshot introduced Kimi K3.1 Pro, a 3T-class open-weight model with a 1-million-token context window, plus a Highspeed variant. Both are available on Kimi.ai, Kimi Work, Kimi Code, and the Kimi API, with full weights due by Nov 4.
Key numbers: Terminal-Bench 4.0: 65.7% vs GPT-6 Astra 58.2% and Claude Opus 5.5 66.4%. Kimi Code Bench 3.0 Long-horizon: 61.2% vs K3's 55.95%.
Why it matters: The company says K3.1 Pro matches or approaches the strongest proprietary models on coding, agentic, and knowledge benchmarks, and it reports gains over K3 across internal finance benchmarks.
Coverage: 4 outlets in 24h
What changed
Moonshot introduced Kimi K3.1 Pro, a 3T-class open-weight model with a 1-million-token context window, plus a Highspeed variant. Both are available on Kimi.ai, Kimi Work, Kimi Code, and the Kimi API, with full weights due by Nov 4.
Why it matters
The company says K3.1 Pro matches or approaches the strongest proprietary models on coding, agentic, and knowledge benchmarks, and it reports gains over K3 across internal finance benchmarks.