In the release page: “The other model(s) in Qwen3.8-series would be released later”, we know that Qwen3.8-27B is dropping next, but that statement implies there might be other models beyond the 27B!
probably would be better for the LLM sub . becase i think subscribers of locallama want to run models on their own hardware.
Can someone give me an ELI5 what’s the difference between 27B and this 2.4T? Is it mostly size? I have a Framework Desktop which has been running 3.5-Coder Next for a while, not sure if I should switch
2.4t is useless unless you’re an institution, you can think of each 1b being approximately 1GB of ram/vram needed just to run it.
2.4T is 4.9TB in size, no way to run that on a Framework Desktop.
27B releases on Friday but should be about 55GB like the old one, that one fits easily and should beat your 3.5-Coder.
Qwen 3.5 122b should also fit easily on a Framework Desktop when quantized and has more knowledge.
Thanks. So you’re saying Qwen 3.5-122b has more knowledge than 3.8-27B? Did I get this right?
Yes. Generally, the larger the model, the more knowledge they have.
However, knowledge is not everything, 27b (even 3.6) is much better than Qwen 3.5 at calling tools and general programming.
27b is also a dense model (27b parameters are active at the same time) while 122b is a MoE model with just 10b active parameters. Usually this makes 27b much smarter at understanding complex situations.
I would recommend trying both and seeing which performs better in your tasks. You can also use something like llama-swap to seamlessly swap between them.
If you need something fast, Qwen3.6-35b is also MoE with 3b active parameters. It is really fast and fairly competent in many fields.
Gemma-4-31b is another dense model that does really well with languages (translations, summaries, …). There’s not really a one size fits all with local models but with swapping them around you can get some great results, especially on a Framework Desktop where multiple of the smaller ones might fit at the same time.
You can also use something like llama-swap to seamlessly swap between them.
FYI: llama-server has built-in model swapping now so you don’t need a second piece of software (at least for the simple cases). I pass
--model-presetsand a path to an INI file with the per-model parameters when I run the server.Thanks, my knowledge on the topic of Local AI is still very barebones. I generally use my FW for coding assistance, troubleshooting Linux issues, automating stuff and doing some heavy text processing (fixing artifacts after OCRing old PDFs and formatting text). I have lots of spare RAM as I’m lucky enough to have the 128GB model I bought before the prices went crazy. I need to try this Qwen3.6 and Gemma as you recommended as it looks like they may fit my needs as well. The Coder I have has been running smoothly for a while and I didn’t really have the opportunity to tinker with other options recently. If you have any tips for my use case I’m all ears!
Maybe if I quantize this down to 0.1-bit it will run on my 3090 and every so often spit out a sentence that isn’t gibberish or random characters…
~2 days until 27B release apparently.
The practical one for local use.
