How to run Qwen 3.5 locally

(unsloth.ai)

57 points | by Curiositry 8 hours ago ago

11 comments

  • moqizhengz an hour ago

    Running 3.5 9B on my ASUS 5070ti 16G with lm studio gives a stable ~100 tok/s. This outperforms the majority of online llm services and the actual quality of output matches the benchmark. This model is really something, first time ever having usable model on consumer-grade hardware.

    • throwdbaaway 13 minutes ago

      There are Qwen3.5 27B quants in the range of 4 bits per weight, which fits into 16G of VRAM. The quality is comparable to Sonnet 4.0 from summer 2025. Inference speed is very good with ik_llama.cpp, and still decent with mainline llama.cpp.

    • lukan 15 minutes ago

      What exact model are you using?

      I have a 16GB GPU as well, but have never run a local model so far. According to the table in the article, 9B and 8-bit -> 13 GB and 27B and 3-bit seem to fit inside the memory. Or is there more space required for context etc?

    • yangikan an hour ago

      Do you point claude code to this? The orchestration seems to be very important.

  • Curiositry an hour ago

    Qwen3.5 9b seems to be fairly competent at OCR and text formatting cleanup running in llama.cpp on CPU, albeit slow. However, I have compiled it umpteen ways and still haven't gotten GPU offloading working properly (which I had with Ollama), on an old 1650 Ti with 4GB VRAM (it tries to allocate too much memory).

    • acters 40 minutes ago

      I have a 1660ti and the cachyos + aur/llama.cpp-cuda package is working fine for me. With about 5.3 GB of usable memory, I find that the 35B model is by far the most capable one that performs just as fast as the 4B model that fits entirely on my GPU. I did try the 9B model and was surprisingly capable. However 35B still better in some of my own anecdotal test cases. Very happy with the improvement. However, I notice that qwen 3.5 is about half the speed of qwen 3

  • Twirrim 4 hours ago

    I've been finding it very practical to run the 35B-A3B model on an 8GB RTX 3050, it's pretty responsive and doing a good job of the coding tasks I've thrown at it. I need to grab the freshly updated models, the older one seems to occasionally get stuck in a loop with tool use, which they suggest they've fixed.

    • fy20 an hour ago

      I guess you are doing offloading to system RAM? What tokens per second do you get? I've got an old gaming laptop with a RTX 3060, sounds like it could work well as a local inference server.

    • ufish235 3 hours ago

      Can you give an example of some coding tasks? I had no idea local was that good.

      • hooch an hour ago

        Changed into a directory recently and fired up the qwen code CLI and gave it two prompts: "so what's this then?" - to which it had a good summary across stack and product, and then "think you can find something todo in the TODO?" - and while I was busy in Claude Code on another project, it neatly finished three HTML & CSS tasks - that I had been procrastinating on for weeks.

        This was a qwen3-coder-next 35B model on M4 Max with 64GB which seems to be 51GB size according to ollama. Have not yet tried the variants from the TFA.

    • fragmede 3 hours ago

      Which models would that be?