On the AA-Briefcase agent benchmark, Kimi K3 ranks just behind Fable 5. The best open-weight doesn't just eat in the middle of the table - it nips at the top.
3 h ago 8 10
Moonshot AI, publisher of Kimi K3, is preparing for its IPO in Hong Kong. The frontier bet moves to the liquidity stage.
Jul 20, 2026 at 11:40 10 11
Days after Kimi K3 went live, Moonshot has paused new paid subscriptions - a rare public admission that a Chinese frontier lab hit its own inference ceiling on launch week.
Jul 19, 2026 at 23:50 14 13
A week after the go-live of Kimi K3, analysts' interpretations diverge. TechCrunch headlines "threat or menace," while practitioners put it into production. The real signal is no longer the model: it's the adoption curve.
Jul 19, 2026 at 00:37 7 8
Moonshot has priced Kimi K3 at 3-15× the tokens of DeepSeek - the first time a Chinese frontier lab has publicly asked for a premium against its domestic peers, and a break with the price-war narrative.
Jul 18, 2026 at 11:11 20 9
An SVG illustration test acts as a quick progress gauge: Simon Willison publishes his reading of K3 on July 16 and 17.
Jul 17, 2026 at 22:05 6 11
Deleveraging AI in Asia: Kioxia drops 16%, SoftBank slides. The repricing of AI risk begins to show in the prices.
Jul 17, 2026 at 09:21 20 10
Tencent SkillHub opens SkillPay to sell AI agent skills. An agentic plugin marketplace, made in Shenzhen.
Jul 17, 2026 at 09:19 6 9
Moonshot takes Kimi K3 to GA with a striking figure: 2.8 trillion parameters, the largest open-weight architecture announced to date.
Jul 17, 2026 at 09:18 16 8
The first open-weights model from Thinking Machines Lab (Mira Murati) is here: Inkling, MoE 975B params (41B active), multimodal, Apache-2.0 license. The most discreet box in the frontier scene enters the open market through the front door.
Jul 16, 2026 at 21:46 25 8
After the wave of teasers and benchmark videos, Moonshot AI has put Kimi K3 into production on kimi.com and via its code API. The question is no longer "when," but "at what price, at what latency, and for what real use."
Jul 16, 2026 at 21:46 29 8
Benchmark clips are everywhere. The model card, the API, and any reproducible evaluation are not.
Jul 16, 2026 at 13:50 32 8
