Kimi

What hackers pick instead

• Qwen 3.8 27BAn open-weights model that runs on a single high-end GPU and nears frontier coding quality, though it overthinks, burns tokens and struggles with complex tasks.1173recommend
• Z.ai GLMAn open-weight coding model close to frontier quality at a fraction of the cost, though critics cite slow, outage-prone hosting, erratic output and expensive self-hosting hardware.1043recommend
• Gemma 4Google's open-weight model family runs well on local hardware with strong quality and token efficiency for its size, though critics cite unreliable tool calling and weaker coding than Qwen.713recommend
• QwenThe go-to open-weight model family for running capable coding and general AI locally on accessible hardware, though it falls short of frontier models like Claude Opus.474recommend
• Qwen3.6-35B-A3BA fast mixture-of-experts open model that runs well on consumer hardware and is surprisingly capable for its size, though it falls well short of frontier models on complex tasks.373recommend

Every quote

The 40 most detailed of 2034 comments. Each links to the original.

1. IMO Kimi K2.5 beats everything but Opus 4.6 and Gemini 3.1... It's much better at most writing tasks, and its "Deep Research" mode is by a wide margin the best in the business.

   Recommends · Feb 2026 · in “Nvidia and OpenAI abandon unfinished $100B deal in favour of $30B investment”

2. It should be clear looking at this model that if adversarial distillation from the closed frontier models in the U.S. contributed, it is at most to a relatively small degree.

   Mixed · Jul 2026 · in “Who's Afraid of Chinese Models?”

3. It is also possible to test it yourself to see how good LLMs are at finding software issues: Find any open source project that isn't huge, then have Kimi 3 review it

   Recommends · Aug 2026 · in “Critical CVE issued for hallucinated SQLite vulnerability”

4. Moonshot has been frontier since this summer. At this point whenever they release a model it's on par with whatever the latest OpenAI/Anthropic model release was.

   Recommends · Oct 2026 · in “Kimi K3.1 Pro and K3.1 Pro Highspeed”

5. other providers can offer third-party open source models with roughly the similar quality like this, Kimi K2.6 or GLM 5.1 for 10 times less the price

   Mixed · Apr 2026 · in “DeepSeek v4”

6. several Chinese models (at least Qwen, DeepSeek, Kimi and GLM) are "close enough" that with the right harness they are cost effective alternatives

   Recommends · Jun 2026 · in “Uber's $1,500/Month AI Limit Is a Useful Signal for AI Tool Pricing”

7. it's the first time I can use an open-source model and not immediately tell the difference between it and top-end models from Anthropic and OpenAI

   Recommends · Jan 2026 · in “Kimi K2.5 Technical Report [pdf]”

8. They did exceed the expecations of pretty much everyone! I've also blogged about using the model, it's a bit on the slow side but pretty good!

   Recommends · Jul 2026 · in “Kimi K3-256k”

9. I have been using a quorum composed of step-3.5-flash, Kimi k2.5 and glm-5 and I have found it outperforms opus-4.5 at a fraction of the cost

   Recommends · Feb 2026 · in “Nvidia and OpenAI abandon unfinished $100B deal in favour of $30B investment”

10. Broadly, I'd say it's quite capable, it even did a ton of validating, took screenshots, identified its own issues, and iterated repeatedly.

    Recommends · Jul 2026 · in “Show HN: Explore 6048 YC companies as an interactive galaxy”

11. the latest versions of Kimi, MiMo, and GLM (pre-5.2) had already surpassed Sonnet in terms of output quality for a fraction of the price

    Recommends · Jun 2026 · in “GLM-5.2 is the new leading open weights model on Artificial Analysis”

12. Chinese open source models are dirt cheap, you can buy $20 worth of kimi-k2.5 on opencode and spam it all week and barely make a dent.

    Recommends · Feb 2026 · in “Why AI Velocity Is Becoming a Debt Accelerator”

13. you may as well just pay for something like Kimi K2.6 via openrouter if you want a something ~>= sonnet 4.6 in performance as a backup

    Recommends · May 2026 · in “Usage-based pricing killing your vibe, here's how to roll your own local AI”

14. Open weight models are available and work great, literally today kimi K3.1 pro was announced and goes toe to toe with closed models.

    Recommends · Oct 2026 · in “New AI Is More Than Twice as Good at Diagnosing Diseases as Doctors”

15. it means neither of the two best financed labs in the world manage to produce even two models themselves that would beat Kimi K3

    Recommends · Jul 2026 · in “Kimi K3 is now live”

16. It's a lot stronger for geospatial intelligence tasks than any other model in my experience. Shame it's so slow in terms of tps

    Mixed · Dec 2025 · in “Kimi K2 1T model runs on 2 512GB M3 Ultras”

17. Personally, I'd rank Kimi K3 above Opus 4.8 and lower than GPT 5.6 Sol in its ability to find vulnerabilities and exploit them.

    Recommends · Jul 2026 · in “UK AISI / Caisi Preliminary Assessment of Kimi K3's Cyber Capabilities”

18. Moonshot open sourcing Kimi K3, a frontier AI and other components really means we are getting abundant AI for all of humanity.

    Recommends · Jul 2026 · in “Kimi-K3 Technical Report [pdf]”

19. Unless frontier labs can surpass the current models significantly I don't know why I would pay them money instead of hosting K3

    Recommends · Jul 2026 · in “Our position on open-weights models”

20. Other than the top performers (e.g. GLM 5.1, Kimi K2.5, where required hardware is basically unaffordable for a single person)

    Recommends · Apr 2026 · in “Claude Opus 4.7 costs 20–30% more per session”

21. I use that specific model as my primary LLM for coding and diagnostics on projects in Python, C++, CUDA, and for web research.

    Recommends · Oct 2026 · in “Kimi K3.1 scores 55 on Artificial Analysis, placed on the Pareto line”

22. Models in the 700B+ category (GLM5, Kimi K2.5) are decent, but running those on your own hardware is a six-figure investment.

    Mixed · Mar 2026 · in “Elevated Errors in Claude.ai”

23. it's been able to write concurrent Go code and debug issues with goroutines equal to, and much more complex then, your issue

    Recommends · Feb 2026 · in “Gemini 3 Deep Think”

24. For HTML/CSS/JavaScript it seems to do fine. Maybe I am just asking too much of it, with a codebase spanning Rust and Elixir

    Mixed · Oct 2026 · in “Kimi K3.1 scores 55 on Artificial Analysis, placed on the Pareto line”

25. I have personally found the prior Kimi K2 Thinking model to be overall more usable and reliable than Gemini 3 Pro and Flash

    Recommends · Jan 2026 · in “OpenAI's Unit Economics”

26. It does small and medium things OK. But if I work on complex things, I'll later have to spend two days cleaning up the mess

    Mixed · Apr 2026 · in “Kimi K2.6: Advancing Open-Source Coding”

27. Kimi costs 0.3/$1.72 on OpenRouter, $200 for that gives you way more than you would get out of a $200 Claude subscription.

    Recommends · Apr 2026 · in “Opus 4.7 to 4.6 Inflation is ~45%”

28. I've been using Kimi K2T and K2.5 to work I'm Go with a fair amount of concurrency

    Recommends · Feb 2026 · in “Gemini 3 Deep Think”

29. Kimi K2.5 is rather competitive in regard to pure output quality

    Recommends · Jan 2026 · in “OpenAI's Unit Economics”

30. my experience with older Kimi models was exactly this, that they'd spin for a whole chunk of time with a lot of back and forth thinking

    Criticizes · Jul 2026 · in “Kimi K3 is now live”

31. GLM and Kimi models are "thinkslop" models that reason over their own thinking, which increases cost and decreases speed significantly

    Criticizes · Aug 2026 · in “Qwen3.8-27B”

32. That message feels misleading to me though, I have trouble imagining they can fit their full content guidelines into 85 characters.

    Criticizes · Jul 2026 · in “Kimi K3 is now live”

33. Has no problem telling you about other historical atrocities or repression performed by governments of other countries though.

    Criticizes · Nov 2025 · in “Kimi K2 Thinking, a SOTA open-source trillion-parameter reasoning model”

34. whenever I try to get my agent to write modern Zig based on a C reference it decides to start writing everything in a C style

    Criticizes · May 2026 · in “We should be more tired than the model”

35. It has internal bias too and the first comment mentions that additional censoring runs on top of the model output in the API.

    Criticizes · Apr 2026 · in “Kimi K2.6: Advancing Open-Source Coding”

36. there are two tiers, one matching Moonshot pricing with comparable latency/throughput, and a fast mode at $4.50/M (still cheaper than Opus) with 3x the throughput.

    Recommends · Jul 2026 · in “Kimi-K3 Releases on HuggingFace 6/27”

37. Kimi and QWEN are now moving on to a restricted-usage license, which, although is still better than the proprietary American models, is a step back

    Mixed · Aug 2026 · in “GLM-5.3: Frontier Coding with Emergent Cyber Capabilities”

38. They tick all the boxes I care about – desktop & mobile app, cli – but for the same price I might as well just go for the leading providers.

    Mixed · Apr 2026 · in “Kimi K2.6: Advancing Open-Source Coding”

39. the K2.5 output looks at least marginally better AIUI, the only real problem with it is the snarky initial "That's very interesting" quip

    Mixed · Jan 2026 · in “Kimi K2.5 Technical Report [pdf]”

40. Extremely useful: the LLM tendency to go overboard on formatting nicely balances out the wall-of-text format from historical publications

    Recommends · Apr 2026 · in “Britannica11.org – a structured edition of the 1911 Encyclopædia Britannica”