Kimi Launches K3.1 Pro and Highspeed Variant, Promises Open Weights by November 4

1. 20:46
   Nathan Lambert · X

   Interconnects Podcast #20 Debates Kimi K3.1 Pro and Open-Weight Cybersecurity

   Interconnects Podcast #20 brings its hosts together to debate difficult questions after a turbulent week in AI. They discuss the mixed reaction to Kimi K3.1 Pro’s benchmarks and whether it will be adopted, the risks of open-weight models and cybersecurity, and why cyber experts and AI experts view recent events differently. The episode also asks what might change in the next few months and what could change the hosts’ minds.

1. 15:18
   Latent SpacePicks

   Moonshot launches Kimi K3.1 Pro with weights promised November 4 and faster Highspeed API

   Moonshot launched Kimi K3.1 Pro, a multimodal MoE model with 2.8 trillion total parameters, 104 billion active parameters, and a 1-million-token context window. Artificial Analysis ranked it fourth with a score of 55, while weights remained unpublished and were promised for November 4, 2026. Standard and Highspeed access launched across Kimi products and API, with Moonshot claiming 2–3× faster output at twice the standard price for Highspeed.

2. 01:51
   宝玉 · XPicks

   Moonshot releases Kimi K3.1 Pro across Kimi products and API

   Moonshot released Kimi K3.1 Pro on October 23, with same-day access through Kimi web and App, Kimi Work, Kimi Code, and the API. According to the official announcement quoted by Baoyu, it is the company's most capable open-weight model, featuring 1M-token context, 2.8T total / 104B active params (MoE), and multimodal capabilities. Baoyu reports that full weights will be released by November 4.

3. 01:20
   Nathan Lambert · X

   Kimi K3.1 announced as 3.1 Pro, with weights promised by November 4th

   Nathan Lambert reports that Kimi K3.1 has been announced as 3.1 Pro and says its benchmark scores appear strongest among current open-weight models, conditional on the promised weights release by November 4th. He highlights the Kimi team's stated approach to moderating open-weight cyber capabilities relative to the closed frontier. K3.1 costs less per token than K3, but Lambert notes that higher token usage on some benchmarks could make per-answer costs similar; informal testing remains an open question.

4. 01:20
   Kimi · X

   Meet Kimi K3.1 Pro, our most capable 3T-class model.

   Meet Kimi K3.1 Pro, our most capable 3T-class model.

5. 01:15
   Kimi · XPicks

   Kimi introduces K3.1 Pro, a 3T-class model with open weights planned

   Kimi introduced Kimi K3.1 Pro with 1M-token context, 2.8T total / 104B active params (MoE), and multimodal capabilities, claiming hallucinations are down by 65% in finance work. K3.1 Pro and K3.1 Pro Highspeed (2–3× faster) are now live on http://Kimi.ai, Kimi Work, Kimi Code, and the Kimi API, with open weights planned by Nov 4, 2026. API: http://platform.kimi.ai Tech blog: http://kimi.ai/blog/kimi-k3-1-pro

1. 18:57
   雷锋网

   Report Says Unreleased Kimi K3.1 Excels at 3D Tasks at Lower Cost

   Leifeng.com reports that Kimi K3.1 has appeared in leaked model listings and user tests, though Moonshot AI has not confirmed a release. Testers cited strong results on Three.js and interactive 3D tasks, and one comparison put its cost at about one-fourteenth that of GPT-6 Astra. The article also notes a weakness: an awkward-looking giraffe animation in one test.

Trends compare the same continuously observed participants, which may cover fewer sources than current heat. Point or click to inspect hourly heat, or use left and right arrow keys.