GLM-4.7-Flash

141 points | by scrlk 2 hours ago

31 comments

  • dajonker 34 minutes ago

    Great, I've been experimenting with OpenCode and running local 30B-A3B models on llama.cpp (4 bit) on a 32 GB GPU so there's plenty of VRAM left for 128k context. So far Qwen3-coder gives the me best results. Nemotron 3 Nano is supposed to benchmark better but it doesn't really show for the kind of work I throw at it, mostly "write tests for this and that method which are not covered yet". Will give this a try once someone has quantized it in ~4 bit GGUF.

    Codex is notably higher quality but also has me waiting forever. Hopefully these small models get better and better, not just at benchmarks.

      latchkey 30 minutes ago
        dajonker 16 minutes ago

        Yes I usually run Unsloth models, however you are linking to the big model now, which I can't run on my consumer hardware.

          latchkey 12 minutes ago

          There are a bunch of 4bit quants in the GGUF link and the 0xSero has some smaller stuff too. Might still be too big and you'll need to ungpu poor yourself.

            disiplus 8 minutes ago

            yeah there is no way to run 4.7 on a 32g vram this flash is something that im also waiting to try later tonight

  • vessenes an hour ago

    Looks like solid incremental improvements. The UI oneshot demos are a big improvement over 4.6. Open models continue to lag roughly a year on benchmarks; pretty exciting over the long term. As always, GLM is really big - 355B parameters with 31B active, so it’s a tough one to self-host. It’s a good candidate for a cerebras endpoint in my mind - getting sonnet 4.x (x<5) quality with ultra low latency seems appealing.

      HumanOstrich 23 minutes ago

      I tried Cerebras with GLM-4.7 (not Flash) yesterday using paid API credits ($10). They have rate limits per-minute and it will get eaten up in the first few seconds - then you have to wait the rest of the minute. Every minute, until your task is done. So they're "fast" at 1000 tok/sec - but not really for practical usage. You effectively get about 17 tok/sec with rate limits.

      They also charge for cached tokens, so I burned through $4 for 1 relatively simple coding task - would've cost <$1 using GPT-5.2-Codex or any other model besides Opus and maybe Sonnet that supports caching. And it would've been much faster.

      mckirk an hour ago

      Note that this is the Flash variant, which is only 31B parameters in total.

      And yet, in terms of coding performance (at least as measured by SWE-Bench Verified), it seems to be roughly on par with o3/GPT-5 mini, which would be pretty impressive if it translated to real-world usage, for something you can realistically run at home.

  • dfajgljsldkjag 40 minutes ago

    Interesting they are releasing a tiny (30B) variant, unlike the 4.5-air distill which was 106B parameters. It must be competing with gpt mini and nano models, which personally I have found to be pretty weak. But this could be perfect for local LLM use cases.

    In my ime small tier models are good for simple tasks like translation and trivia answering, but are useless for anything more complex. 70B class and above is where models really start to shine.

  • bilsbie 22 minutes ago

    What’s the significance of this for someone out of the loop?

      epolanski 9 minutes ago

      You can run gpt 5 mini level ai on your MacBook with 32 gb ram.

  • eurekin 28 minutes ago

    I'm trying to run it, but getting odd errors. Has anybody managed to run it and can share the command?

  • karmakaze an hour ago

    Not much info than being a 31B model. Here's info on GLM-4.7[0] in general.

    I suppose Flash is merely a distillation of that. Filed under mildly interesting for now.

    [0] https://z.ai/blog/glm-4.7

      lordofgibbons 42 minutes ago

      How interesting it is depends purely on your use-case. For me this is the perfect size for running fine-tuning experiments.

      redrove 40 minutes ago

      A3.9B MoE apparently

  • twelvechess an hour ago

    Excited to test this out. We need a SOTA 8B model bad though!

  • XCSme an hour ago

    Seems to be marginally better than gpt-20b, but this is 30b?

      strangescript an hour ago

      I find gpt-oss 20b very benchmaxxed and as soon as a solution isn't clear it will hallucinate.

        blurbleblurble 21 minutes ago

        Every time I've tried to actually use gpt-oss 20b it's just gotten stuck in weird feedback loops reminiscent of the time when HAL got shut down back in the year 2001. And these are very simple tests e.g. I try and get it to check today's date from the time tool to get more recent search results from the arxiv tool.

      lostmsu an hour ago

      It actually seems worse. gpt-20b is only 11 GB because it is prequantized in mxfp4. GLM-4.7-Flash is 62 GB. In that sense GLM is closer to and actually is slightly larger than gpt-120b which is 59 GB.

      Also, according to the gpt-oss model card 20b is 60.7 (GLM claims they got 34 for that model) and 120b is 62.7 on SWE-Bench Verified vs GLM reports 59.7

  • epolanski an hour ago

    Any cloud vendor offering this model? I would like to try it.

      PhilippGille an hour ago

      z.ai itself, or Novita fow now, but others will follow soon probably

      https://openrouter.ai/z-ai/glm-4.7-flash/providers

        epolanski an hour ago

        Interesting, it costs less than a tenth than Haiku.

          saratogacx 30 minutes ago

          GLM itself is quite inexpensive. A year sub to their coding plan is only $29 and works with a bunch of various tools. I use it heavily as a "I don't want to spend my anthropic credits" day-to-day model (mostly using Crush)

      dvs13 an hour ago
      latchkey 27 minutes ago

      We don't have lot of GPUs available right now, but it is not crazy hard to get it running on our MI300x. Depending on your quant, you probably want a 4x.

      ssh admin.hotaisle.app

      Yes, this should be made easier to just get a VM with it pre-installed. Working on that.

        omneity 22 minutes ago

        Unless using docker, if vllm is not provided and built against ROCm dependencies it’s going to be time consuming.

        It took me quite some time to figure the magic combination of versions and commits, and to build each dependency successfully to run on an MI325x.

          latchkey 16 minutes ago

          Agreed, the OOB experience kind of suck.

          Here is the magic (assuming a 4x)...

            docker run -it --rm \
            --pull=always \
            --ipc=host \
            --network=host \
            --privileged \
            --cap-add=CAP_SYS_ADMIN \
            --device=/dev/kfd \
            --device=/dev/dri \
            --device=/dev/mem \
            --group-add render \
            --cap-add=SYS_PTRACE \
            --security-opt seccomp=unconfined \
            -v /home/hotaisle:/mnt/data \
            -v /root/.cache:/mnt/model \
            rocm/vllm-dev:nightly
            
            mv /root/.cache /root/.cache.foo
            ln -s /mnt/model /root/.cache
            
            VLLM_ROCM_USE_AITER=1 vllm serve zai-org/GLM-4.7-FP8 \
            --tensor-parallel-size 4 \
            --kv-cache-dtype fp8 \
            --quantization fp8 \
            --enable-auto-tool-choice \
            --tool-call-parser glm47 \
            --reasoning-parser glm45 \
            --load-format fastsafetensors \
            --enable-expert-parallel \
            --allowed-local-media-path / \
            --speculative-config.method mtp \
            --speculative-config.num_speculative_tokens 1 \
            --mm-encoder-tp-mode data
      xena an hour ago

      The model literally came out less than a couple hours ago, it's going to take people a while in order to tool it for their inference platforms.

        idiliv an hour ago

        Sometimes model developers coordinate with inference platforms to time releases in sync.