







Sync from your watch. Bring your coach along. Own every byte, forever.
Login | TrainingPeaks
Log in to your TrainingPeaks athlete edition or coach edition to track, analyze and plan.
Kimi K3 - Kimi API Platform
Kimi K3 is our flagship model for long-horizon coding and end-to-end knowledge work, with a 1M-token context window and industry-leading intelligence. The Kimi API Platform provides K3, K2.7 Code, K2.6 and other large language model APIs, supporting long context, multimodal understanding, and Tool Calling.

Kimsufi New Homepage
Rent an affordable dedicated server, powered by our global infrastructure. Unlimited traffic, built-in anti-DDoS, and Root access. Bare Metal performance for websites and gaming.
Post-Training 50x Faster
We're announcing Trellis, the fastest open-source post-training code for Kimi K2 Thinking

Flagship Model Kimi K3 Pricing - Kimi API Platform
Kimi K3 is our flagship model for long-horizon coding and end-to-end knowledge work, with a 1M-token context window and industry-leading intelligence. The Kimi API Platform provides K3, K2.7 Code, K2.6 and other large language model APIs, supporting long context, multimodal understanding, and Tool Calling.


On Kimi K3: Its Capabilities And Related Discontents
Kimi K3 is a very good model with excellent benchmarks.

Kimi K2: Open Agentic Intelligence
Kimi K2 is our latest Mixture-of-Experts model with 32 billion activated parameters and 1 trillion total parameters. It achieves state-of-the-art performance in frontier knowledge, math, and coding among non-thinking models.
Kimi-K3/k3_tech_report.pdf at main · MoonshotAI/Kimi-K3
Open Frontier Intelligence. Contribute to MoonshotAI/Kimi-K3 development by creating an account on GitHub.
unpingable/atproto-driftwatch
Tracks how claims propagate, mutate, and concentrate across the Bluesky ATProto network.
5 Thoughts on Kimi K2 Thinking
Quick thoughts on another fantastic open model from a rapidly rising Chinese lab.

Kimi K3 Tech Blog: Open Frontier Intelligence
Kimi K3 is the world's first open 3T-class model — frontier performance across coding, knowledge work, and reasoning, with native multimodality and 1M context.
Library — cameron.stream
How LoRA matches full training performance more broadly than expected.
Malfestio — Learning on the AT Protocol
Coming soon: Master complex topics with spaced repetition, linked notes, and active recall. Share your decks, notes, and discoveries with the community.

Zhuokai Zhao on Twitter / X
Tons of interesting things in the Kimi K3 tech report — here are five algorithm-side techniques that I think either I've never seen before or simply deserve more attention than they're getting.1/ They open-sourced the model but kept the speculative decoding draft model, which…— Zhuokai Zhao (@zhuokaiz) July 30, 2026
Kuba Suder 🇵🇱🇺🇦
@mackuba.eu
Indie Mac/iOS & web developer • Currently building random fun things on ATProto using Ruby and occasionally JavaScript → blue.mackuba.eu "mostly-happy-but-occasionally-grumpy Ruby guy" - Why 🌍 Kraków, Poland 🇵🇱🇪🇺 • ✝️