upvote

points

by throwaw121 hours ago |
comments
upvoteby 2ndorderthought1 hours ago|
next
[-]
It's pretty close already. Check qwen3.6 27b if you haven't already. People are vibe and agentic coding with it on a single GPU.

It is more finicky than Claude but if you hand hold it a bit it's crazy.

reply
upvoteby pizza23412 minutes ago|
parent|
next
[-]
Vibe coding on consumer hardware is still very limited; this is especially true on GPUs, whose RAM limit is around 16 - maybe 24 - GB for the vast majority (although Macs change the equation).

These are two realworld experiments, whose results are disappointing for those expecting levels of performance comparable to cloud services:

- https://deploy.live/blog/running-local-llms-offline-on-a-ten...

- https://betweentheprompts.com/40000-feet/

The first is even the 35b version of qwen3.6.

reply
upvoteby 2ndorderthought2 minutes ago|
parent|
[-]
I don't see how it's disappointing? 95% correct using the 35b model before the right quants dropped while most people are preferring 27b on a laptop? And they still got tons of code written for them.

On a real GPU using 27b with the latest quants the experience is better. It's still not the same as opus running on a subsidized GPU farm. Well it is better for privacy at least.

I find it interesting how 2 people can read the same thing and come to very different conclusions.

reply
upvoteby gchamonlive1 hours ago|
parent|
prev|
next
[-]
I see that going around, and either the test cases are too simplistic or I'm doing something wrong. I have a server with a 3090 in it, enough to run qwen3.6, but I haven't had much luck using it with either codex or oh-my-pi. They work, but the model gets really slow with ~64k context and the attention degrades quickly. You'll sometimes execute a prompt, the model will load a test file and say something like "I was presented with a test file but no command. What should I do with it?".

So yeah, while it's true that qwen3.6 is good for agentic coding, it's not very good for exploring the codebase and coming up with plans. You need to pair it today with a model capable of ingesting the whole context and providing a detailed plan, and even then the implementation might take 10x the amount of time it'd take for sonnet or Gemini 3 to crunch through the plan.

EDIT:

My setup is really as simple as possible. I run ollama on a remote server on my local network. In my laptop I set OLLAMA_HOST and do `ollama pull qwen3.6:27b`, which then becomes available to the agent harnesses. I am not sure now how I set the context, but I think it was directly in oh-my-pi. So server config- and quantization-wise, it's the defaults.

reply
upvoteby dminik42 minutes ago|
parent|
next
[-]
Yeah. Context size matters a lot. With OpenCode dumping like 10k tokens in the system prompt it takes like 4 rounds before it had to compact at say 64k. It's not really worth it to run at anything below 100k and even then the models aren't all that useful.

They're also pretty terrible at summarization. Pretty much always some file read or write in the middle of the task would cross the context margin and it would mark it as completed in the summary. I think leaving the first prompt as well as the last few turns intact would improve this issue quite a lot, but at low context sizes thats pretty much the whole context ...

reply
upvoteby pferdone52 minutes ago|
parent|
prev|
next
[-]
I can see that and I don't know your setup, but there are people pushing >70t/s with MTP on a single 3090, with big contexts still >50t/s. 64k is not a lot for agentic coding, and IIRC 128k with turboquant and the likes should be possible for you. r/LocalLLM/ and r/LocalLLaMA/ are worth a visit IMO.

EDIT: just found this recipe repo, may wanna give it a go: https://github.com/noonghunna/club-3090

EDIT-2: this can also shave off a lot of context need for tool calling -> https://github.com/rtk-ai/rtk

reply
upvoteby gchamonlive18 minutes ago|
parent|
[-]
will give more info in the post

EDIT: thanks for the links!

reply
upvoteby embedding-shape59 minutes ago|
parent|
prev|
next
[-]
You're not sharing what quantization you're using, in my experience, anything below Q8 and less than ~30B tends to basically be useless locally, at least for what you typically use codex et al for, I'm sure it works for very simple prompts.

But as soon as you go below Q8, the models get stuck in repeating loops, get the tool calling syntax wrong or just starts outputting gibberish after a short while.

reply
upvoteby gchamonlive19 minutes ago|
parent|
[-]
will do that in an edit to the post
reply
upvoteby embedding-shape3 minutes ago|
parent|
[-]
Sure, waiting :)

In the meantime, Ollama seems to default to "Q4_K_M" which is barely usable for anything, and really won't be useful for agentic coding, the quantization level is just too low. Not sure why Ollama defaults to basically unusable quantizations, but that train left a long time ago, they're more interesting in people thinking they can run stuff, rather than flagging things up front, and been since day 1.

reply
upvoteby nixon_why6926 minutes ago|
parent|
prev|
next
[-]
Qwen3.6 supports 266k context out of the box. Try using q8 kv cache to enable more of it.
reply
upvoteby gchamonlive13 minutes ago|
parent|
[-]
I limited it to 64k expecting 24GB vram to not be enough to make use of the entire context window, but I'll try with other's suggestions.
reply
upvoteby 2ndorderthought1 hours ago|
parent|
prev|
next
[-]
I agree for planning it's not there yet. But I wouldn't be surprised if something came out that was in a similar weight class.
reply
upvoteby regexorcist1 hours ago|
parent|
prev|
[-]
Try oh-my-openagent plan mode.
reply
upvoteby iugtmkbdfil83421 minutes ago|
parent|
prev|
[-]
Eh. It is good in terms of results ( accuracy, good recommendations and so on ), but slow when it comes to actual inference. On local 128gb machine, it took over 5 minutes to brainstorm garage door opening mechanism with some additional restrictions for spice.
reply
upvoteby steveharing11 hours ago|
prev|
[-]
That's absolutely possible, its just as we move towards more advancement, We'll soon see Small models being smart enough to not be judged by parameter count but their reasoning and intelligence. You can see examples like Qwen 3.6 27B.
reply
upvoteby regexorcist1 hours ago|
parent|
[-]
Yeah this is key, a lot of people are still just looking at the number of params and thinking these models are toys. What Qwen 3.6 has shown is that reasoning and tool calling are just as important if not more.
reply