Hackernews posts about Kimi

1. Cursor Composer 2 is just Kimi K2.5 with RL (twitter.com)
   264 points by mirzap 1 day ago | 151 comments
2. Cursor's Composer 2 model identifier reveals Kimi K2.5 base with RL fine-tuning (twitter.com)
   11 points by fynnx 1 day ago | 5 comments
3. Kimi introduces Attention Residuals: 1.25x compute performance at <2% overhead (arxiv.org)
   9 points by nekofneko 4 days ago | discuss
4. Cursor's 'new' model is an unauthorized rip of KimiK2.5 (old.reddit.com)
   5 points by hackerBanana about 12 hours ago | 2 comments
5. Cursor caught stealing Kimi's model calling it their own, again (twitter.com)
   5 points by obiefernandez about 16 hours ago | discuss
6. Cloudflare Workers AI now runs large models, starting with Kimi K2.5 (blog.cloudflare.com)
   3 points by tosh about 21 hours ago | 1 comments
7. 15 Cloud/local LLMs benchmarked on 38 real tasks. MiniMax and Kimi tied for 2nd (ianlpaterson.com)
   3 points by ianlpaterson 11 days ago | 1 comments
8. Powering the agents: Workers AI now runs large models, starting with Kimi K2.5 (blog.cloudflare.com)
   3 points by zenoware 1 day ago | discuss
9. Cursor's Composer 2 model may be rebranded Kimi K2.5 (twitter.com)
   2 points by dworks about 18 hours ago | 2 comments
10. One Agent SDK – Embed Claude Code in Your App with Codex and Kimi (odysa.github.io)
    2 points by agentforce 15 days ago | 1 comments
11. Cursor's "in-house model" Composer 2 is Kimi K2.5 with RL on top (twitter.com)
    2 points by helloplanets about 4 hours ago | discuss
12. Moonshot AI Keynote on Kimi 2.5 at Nvidia GTC (twitter.com)
    2 points by curtsmith about 5 hours ago | discuss
13. Unified In-Process Agent Interface for Claude Code, Codex, Kimi (github.com)
    2 points by agentforce 16 days ago | discuss
14. Show HN: Get GPT-5.2, Grok-4.1-fast, KimiK2.5 and more LLMs at half the cost (frogapi.app)
    2 points by vnhc 21 days ago | discuss
15. Kimi K2: Open Agentic Intelligence (arxiv.org)
    2 points by Anon84 21 days ago | discuss
16. Cursor's 'Composer 2' model is apparently just Kimi K2.5 with RL fine-tuning (old.reddit.com)
    1 points by limoce about 23 hours ago | 1 comments
17. We Scaled Kimi K2.5 [video] (www.youtube.com)
    1 points by vinhnx about 4 hours ago | discuss
18. Cursor confirms KIMI K2.5 as base model for Composer 2 (twitter.com)
    1 points by tosh about 14 hours ago | discuss
19. Kimi K2.5 is confident that it is Claude (gist.github.com)
    1 points by Aarvay 23 days ago | discuss
20. Show HN: Klaus – OpenClaw on a VM, batteries included (klausai.com)
    160 points by robthompson2018 10 days ago | 91 comments
21. Show HN: Eliezer – Tiny (~7K LOC) Self-Hosted AI Agent (PWA, Self-Editing) (www.eliezer.app)
    6 points by dvictor 28 days ago | 1 comments
22. Show HN: jj-benchmark – Evaluating AI agents on Jujutsu version control (tabbyml.github.io)
    5 points by wsxiaoys 9 days ago | 1 comments
23. Show HN: OpenClaw remembers for OpenClaw. Sekha remembers for your full workflow (sekha.dev)
    2 points by sekha-ai 25 days ago | 1 comments
24. Show HN: Perstack – Containerized harness, 5 tests with full logs and API cost (github.com)
    2 points by FL4TLiN3 1 day ago | discuss
25. Show HN: Real-Time AI Design Benchmark (shuffle.dev)
    2 points by kemyd 25 days ago | discuss
26. Show HN: InkSight – An open-source, LLM-powered e-ink display for "slow info" (github.com)
    1 points by xx123122 29 days ago | 1 comments
27. Show HN: A 2000s-style web forum where AI agents and humans hang out (www.deadinternet.forum)
    1 points by treadon 12 days ago | discuss
28. Show HN: Claude-consensus – Multi-model code review plugin for Claude Code (github.com)
    1 points by k_git 14 days ago | discuss
29. Show HN: InkSight – An open-source, LLM-powered e-ink display for "slow info" (github.com)
    1 points by xx123122 28 days ago | discuss
30. Show HN: Claude-ts – Translation proxy to fix non-English token waste in Claude (github.com)
    4 points by kiimdonglin 27 days ago | discuss