[HN Gopher] Qwen3.5 122B and 35B models offer Sonnet 4.5 perform...
       ___________________________________________________________________
        
       Qwen3.5 122B and 35B models offer Sonnet 4.5 performance on local
       computers
        
       Author : lostmsu
       Score  : 97 points
       Date   : 2026-02-28 20:20 UTC (2 hours ago)
        
 (HTM) web link (venturebeat.com)
 (TXT) w3m dump (venturebeat.com)
        
       | xenospn wrote:
       | Are there any non-Chinese open models that offer comparable
       | performance?
        
         | MarsIronPI wrote:
         | I think you could look into Minstral. There's also GPT-OSS but
         | I'm not sure how well it stacks up.
         | 
         | What's your problem with Chinese LLMs?
        
       | aliljet wrote:
       | Is this actually true? I want to see actual evals that match this
       | up with Sonnet 4.5.
        
         | lostmsu wrote:
         | Not exactly, but pretty close:
         | https://artificialanalysis.ai/models/capabilities/coding?mod...
         | 
         | Somewhere between Haiku 4.5 and Sonnet 4.5
        
           | CharlesW wrote:
           | > _Somewhere between Haiku 4.5 and Sonnet 4.5_
           | 
           | That's like saying "somewhere between Eliza and Haiku 4.5".
           | Haiku is not even a so-called 'reasoning model'.1
           | 
           | 1 _To preempt the easily-offended, this is what the latest
           | Opus 4.6 in today 's Claude Code update says: "Claude Haiku
           | 4.5 is not a reasoning model -- it's optimized for speed and
           | cost efficiency. It's the fastest model in the Claude family,
           | good for quick, straightforward tasks, but it doesn't have
           | extended thinking/reasoning capabilities."_
        
             | pityJuke wrote:
             | Haiku 4.5 is a reasoning model. [0]
             | 
             | [0]: https://www-
             | cdn.anthropic.com/7aad69bf12627d42234e01ee7c3630...
             | 
             | > Claude Haiku 4.5, a new hybrid reasoning large language
             | model from Anthropic in our small, fast model class.
             | 
             | > As with each model released by Anthropic beginning with
             | Claude Sonnet 3.7, Claude Haiku 4.5 is a hybrid reasoning
             | model. This means that by default the model will answer a
             | query rapidly, but users have the option to toggle on
             | "extended thinking mode", where the model will spend more
             | time considering its response before it answers. Note that
             | our previous model in the Haiku small-model class, Claude
             | Haiku 3.5, did not have an extended thinking mode.
        
               | CharlesW wrote:
               | Sure, marketing people gonna market. But Haiku's
               | 'extended thinking' mode is very different than the
               | reasoning capabilities of Sonnet or Opus.
               | 
               | I would absolutely believe mar-ticles that Qwen has
               | achieved Haiku 4.5 'extended thinking' levels of coding
               | prowess.
        
               | DetroitThrow wrote:
               | >Sure, marketing people gonna market.
               | 
               | Oh HN never change.
        
               | CharlesW wrote:
               | I'm marketing people, I can say that.
        
           | pinum wrote:
           | Looks much closer to Haiku than Sonnet.
           | 
           | Maybe "Qwen3.5 122B offers Haiku 4.5 performance on local
           | computers" would be a more realistic and defensible claim.
        
         | magicalhippo wrote:
         | The Qwen3.5 27B model did almost the same as Sonnet 4.5 in
         | this[1] reasoning benchmark, results here[2].
         | 
         | Obviously there's more to a model than that but it's a data
         | point.
         | 
         | [1]: https://github.com/fairydreaming/lineage-bench
         | 
         | [2]: https://github.com/fairydreaming/lineage-bench-
         | results/tree/...
        
       | mark_l_watson wrote:
       | The new 35b model is great. That said, it has slight
       | incompatibility's with Claude Code. It is very good for tool use.
        
         | johnnyApplePRNG wrote:
         | Claude code is designed for anthropic models. Try it with
         | opencode!
        
           | kristianpaul wrote:
           | Or Pi
        
             | copperx wrote:
             | Or Oh My Pi
        
       | erelong wrote:
       | What kind of hardware does HN recommend or like to run these
       | models?
        
         | xienze wrote:
         | It's less than you'd think. I'm using the 35B-A3B model on an
         | A5000, which is something like a slightly faster 3080 with 24GB
         | VRAM. I'm able to fit the entire Q4 model in memory with 128K
         | context (and I think I would probably be able to do 256K since
         | I still have like 4GB of VRAM free). The prompt processing is
         | something like 1K tokens/second and generates around 100
         | tokens/second. Plenty fast for agentic use via Opencode.
        
           | rahimnathwani wrote:
           | There seem to be a lot of different Q4s of this model:
           | https://www.reddit.com/r/LocalLLaMA/s/kHUnFWZXom
           | 
           | I'm curious which one you're using.
        
             | suprjami wrote:
             | Unsloth Dynamic. Don't bother with anything else.
        
               | rahimnathwani wrote:
               | UD-Q4_K_XL?
        
           | msuniverse2026 wrote:
           | I've had an AMD card for the last 5 years, so I kinda just
           | tuned out of local LLM releases because AMD seemed to abandon
           | rocm for my card (6900xt) - Is AMD capable of anything these
           | days?
        
             | wirybeige wrote:
             | The vulkan backend for llama.cpp isn't that far behind rocm
             | for pp and tp speeds
        
         | suprjami wrote:
         | The cheapest option is two 3060 12G cards. You'll be able to
         | fit the Q4 of the 27B or 35B with an okay context window.
         | 
         | If you want to spend twice as much for more speed, get a
         | 3090/4090/5090.
         | 
         | If you want long context, get two of them.
         | 
         | If you have enough spare cash to buy a car, get an RTX Ada with
         | 96G VRAM.
        
           | barrkel wrote:
           | Rtx 6000 pro Blackwell, not ada, for 96GB.
        
         | dajonker wrote:
         | Radeon R9700 with 32 GB VRAM is relatively affordable for the
         | amount of RAM and with llama.cpp it runs fast enough for most
         | things. These are workstation cards with blower fans and they
         | are LOUD. Otherwise if you have the money to burn get a 5090
         | for speeeed and relatively low noise, especially if you limit
         | power usage.
        
         | andsoitis wrote:
         | For fast inference, you'd be hard pressed to beat an Nvidia RTX
         | 5090 GPU.
         | 
         | Check out the HP Omen 45L Max: https://www.hp.com/us-
         | en/shop/pdp/omen-max-45l-gaming-dt-gt2...
        
           | laweijfmvo wrote:
           | I never would have guessed that in 2026, data centers would
           | be measured in Watts and desktop PCs measured in liters.
        
         | CamperBob2 wrote:
         | I think the 27B dense model at full precision and 122B MoE at
         | 4- or 6-bit quantization are legitimate killer apps for the 96
         | GB RTX 6000 Pro Blackwell, if the budget supports it.
         | 
         | I imagine any 24 GB card can run the lower quants at a
         | reasonable rate, though, and those are still very good models.
         | 
         | Big fan of Qwen 3.5. It actually delivers on some of the hype
         | that the previous wave of open models never lived up to.
        
           | MarsIronPI wrote:
           | I've had good experience with GLM-4.7 and GLM-5.0. How would
           | you compare them with Qwen 3.5? (If you have any experience
           | with them.)
        
         | zozbot234 wrote:
         | It depends. How much are you willing to wait for an answer?
         | Also, how far are you willing to push quantization, given the
         | risk of degraded answers at more extreme quantization levels?
        
         | elorant wrote:
         | Macs or a strix halo. Unless you want to go lower than 8-bit
         | quantization where any GPU with 24GBs of VRAM would probably
         | run it.
        
       | solarkraft wrote:
       | What are the recommended 4 bit quants for the 35B model? I don't
       | see official ones:
       | https://huggingface.co/models?other=base_model:quantized:Qwe...
       | 
       | Edit: The unsloth quants seem to have been fixed, so they are
       | probably the go-to again:
       | https://unsloth.ai/docs/models/qwen3.5/gguf-benchmarks
        
       | solarkraft wrote:
       | Smells like hyperbole. A lot of people making such claims don't
       | seem to have continued real world experience with these models or
       | seem to have very weird standards for what they consider usable.
       | 
       | Up until relatively recently, while people had already long been
       | making these claims, it came with the asterisks of ,,oh, but you
       | can't practically use more than a few K tokens of context".
        
         | tempest_ wrote:
         | Qwen3-Coder-30B-A3B-Instruct is good I think for in line IDE
         | integration or operating on small functions or library code but
         | I dont think you will get too far with one shot feature
         | implementation that people are currently doing with Claude or
         | whatever.
        
           | andy_ppp wrote:
           | I have been adding a one shot feature to a codebase with
           | ChatGPT 5.3 Codex in Cursor and it worked out of the box but
           | then I realised everything it had done was super weird and it
           | didn't work under a load of edge cases. I've tried being
           | super clear about how to fix it but the model is lost. This
           | was not a complex feature at all so hopefully I'm employed
           | for a few more years yet.
        
       | kristianpaul wrote:
       | They work great with kagi and pi
        
       | alexpotato wrote:
       | I recently wrote a guide on getting:
       | 
       | - llama.cpp
       | 
       | - OpenCode
       | 
       | - Qwen3-Coder-30B-A3B-Instruct in GGUF format (Q4_K_M
       | quantization)
       | 
       | working on a M1 MacBook Pro (e.g. using brew).
       | 
       | It was bit finicky to get all of the pieces together so hopefully
       | this can be used with these newer models.
       | 
       | https://gist.github.com/alexpotato/5b76989c24593962898294038...
        
         | copperx wrote:
         | How fast does it run on your M1?
        
         | robby_w_g wrote:
         | Does your MBP have 32 GB of ram? I'm waiting on a local model
         | that can run decently on 16 GB
        
         | freeone3000 wrote:
         | We can also run LM Studio and get it installed with one search
         | and one click, exposed through an OpenAI-compatible API.
        
       | sunkeeh wrote:
       | Qwen3.5-122B-A10B BF16 GGUF = 224GB. The "80Gb VRAM" mentioned
       | here will barely fit Q4_K_S (70GB), which will NOT perform as
       | shown on benchmarks.
       | 
       | Quite misleading, really.
        
       | kristianpaul wrote:
       | https://unsloth.ai/docs/models/qwen3.5#qwen3.5-27b " Qwen3.5-27B
       | For this guide we will be utilizing Dynamic 4-bit which works
       | great on a 18GB RAM"
        
         | kristianp wrote:
         | 18GB was an odd 3-channel one-off for the M3 Pros. I guess
         | there's a bunch of them out there, but how slow would 27B be on
         | it, due to not being an MOE model.
        
       | gunalx wrote:
       | qwen 3.5 is really decent. oOtside for some weird failures on
       | some scaffolding with seemingly different trained tools.
       | 
       | Strong vision and reasoning performance, and the 35-a3b model run
       | s pretty ok on a 16gb GPU with some CPU layers.
        
       | PunchyHamster wrote:
       | I asked it to recite potato 100 times coz I wanted to benchmark
       | speed of CPU vs GPU. It's on 150 line of planning. It recited the
       | requested thing 4 times already and started drafting the 5th
       | response.
       | 
       | ...yeah I doubt it
        
         | lumirth wrote:
         | well hold on now, maybe it's onto something. do you really know
         | what it means to "recite" "potato" "100" "times"? each of those
         | words could be pulled apart into a dissertation-level thesis
         | and analysis of language, history, and communication.
         | 
         | either that, or it has a delusional level of instruction
         | following. doesn't mean it can't code like sonnet though
        
       | mstaoru wrote:
       | I periodically try to run these models on my MBP M3 Max 128G
       | (which I bought with a mind to run local AI). I have a certain
       | deep research question (in a field that is deeply familiar to me)
       | that I ask when I want to gauge model's knowledge.
       | 
       | So far Opus 4.6 and Gemini Pro are very satisfactory, producing
       | great answers fairly fast. Gemini is very fast at 30-50 sec, Opus
       | is very detailed and comes at about 2-3 minutes.
       | 
       | Today I ran the question against local qwen3.5:35b-a3b - it
       | puffed for 45 (!) minutes, produced a very generic answer with
       | errors, and made my laptop sound like it's going to take off any
       | moment.
       | 
       | Wonder what am I doing wrong?..
        
       ___________________________________________________________________
       (page generated 2026-02-28 23:00 UTC)