4 ms·
GLM-4.7-Flash was the first local coding model that I felt was intelligent enough to be useful. It feels something like Claude 4.5 Haiku at a parameter size whe
by 2001zhaozhao 9mo ago
GLM-4.7-Flash was the first local coding model that I felt was intelligent enough to be useful. It feels something like Claude 4.5 Haiku at a parameter size where other coding models are still getting into loops and making bewilderingly stupid tool calls. It also has very clear reasoning traces that feel like Claude, which does result in the ability to inspect its reasoning to figure out why it made certain decisions.
So far I haven't managed to get comparably good results out of any other local model including Devstral 2 Small and the more recent Qwen-Coder-Next.
- khimaros 9mo agominimax-m.2 is close
- satvikpendem 9mo ago2.5 is out now too.
- khimaros 9mo agoi meant m2.1, but you are probably talking about kimi, not minimax
- satvikpendem 9mo agoNo, MiniMax M2.5 is now available on agent.minimax.io. We await the weights still.
- satvikpendem 9mo agohttps://www.minimax.io/news/minimax-m25 https://www.minimax.io/news/minimax-m25
- Balinares 9mo agoI'm not sure what it is about GLM 4.7 Flash, but it definitely seems to nail a sweet spot. Even the supposedly frontier models make a mess of large requests, so small, well-scoped requests are the way, IMO; and in that space, 4.7 Flash holds its own better than it has any right to.
- Aerroon 9mo agoAnd you can run quantized versions on old hardware! Like 10 year old hardware. You might only get 3 tokens/sec, but it works.
- nake89 9mo agoSlightly off topic. I had a hard time getting models to run with ollama, and I thought that my computer (32gm ram, GTX4070 12Gb vram) just couldn't do it. The I tried LM Studio and after fiddling with some settings, I got models running and quite fast. I didn't try GLM-4.7 flash but I did GLM-4.6v flash and it was amazing to see it be able to analyze all kinds of images (since it has vision support). I was simply stunned. I can't believe that a simple gaming machine can do many of the things I used cloud models for. It was absolutely strikingly good at guessing locations of photos. Even vague ones. Deducing landmarks, writings, types of traffic signs. I need to try 4.7 flash. Hopefully it can ran fast with my machine.
- ThouYS 9mo agofor me gpt-oss:20b was that. glm 4.7 flash was not better, but much slower on a 16GB card