References 684% CONFIDENCEOverall confidence: 84%How well the pin's source and references back up its dates.Weighted average of how firmly 6 references, the source included, support the pin's start and end times; a reference counts half as much for every 180 days older than the newestShow all pins at 75% confidence or better
The first entry is always the pin's source. Overall confidence is a weighted average of how firmly each reference supports the start and end times used above; a reference counts half as much for every 180 days older than the newest.
- [1]90%kimi.com/blog/kimi-k3kimi.com· Posted Oct 12, 2026· Starts Jul 16, 2026 ✓· 21% of score
Moonshot's[2] tech blog opens "Today, we are introducing Kimi K3" and says it is available that day; its benchmark notes are "current as of July 16, 2026", Fortune[5] reported the launch on 16 July 2026 and OpenRouter listed kimi-k3 on 2026-07-16. The weights followed on 27 July, as the launch post promised.
- [2]90%Model Inference Pricing Explanation - Kimi API Platformplatform.moonshot.ai· Added Oct 12, 2026· 21% of score
Moonshot's API pricing page: kimi[1]-k3 at $0.30 per million cached input tokens, $3.00 input, $15.00 output and $3.00 / $6.00 cache writes (5 min / 1 h), with a 1,048,576-token context window.
- [3]90%moonshotai/Kimi-K3 - Hugging Facehuggingface.co· Published Jul 27, 2026· 15% of score
Moonshot's[2] model card: 2.8T total and 104B activated parameters, 93 layers (69 KDA + 24 gated MLA), 16 of 896 experts, MoonViT-V2 vision encoder, 1,048,576-token context, MXFP4 weights and the full max-effort benchmark table; weights made public on 27 July 2026 under the Kimi[1] K3 license.
- [4]75%Moonshot AI releases Kimi K3, a 2.8-trillion-parameter open-weight model rivaling top US systems - MLQmlq.ai· Published Jul 17, 2026· 15% of score
17 July report: about 2.8x the size of K2.6, $3 / $15 per million tokens, no model card or weights at launch, and a reported 27% drop in Z.ai's shares and 16% in MiniMax's after the announcement.[1]
- [5]80%Moonshot's Kimi K3 pushes Chinese AI into Fable-level territory - Fortunefortune.com· Published Jul 16, 2026· 15% of score
Nicholas Gordon's 16 July report: Moonshot's[2] claims against Anthropic's Fable 5, Opus 4.8 and OpenAI's GPT 5.6 Sol, the output-price comparison ($15 against DeepSeek V4's $0.87, GLM-5.2's $4.40 and Fable's $50) and Moonshot's $2 billion May raise at a valuation above $20 billion.
Suggest a correction
Something missing or wrong? Say it in your own words: a link that backs this pin up, a different start or end date and why, or a fact it lacks or gets wrong. The AI checks it against this pin's sources, searches for better ones, and adds any page that backs you up. The pin's own sources still count most. A picture that shows something else, or shows it badly, is looked at too, and moved down or replaced.