Re:Gemma 4 12B: A unified, encoder-free multimodal model
by hedgehog on Wednesday June 03, 2026 @09:18PM
(#48392488)
The 6-bit versions + 8-bit KV cache seems to save a good bit of memory without a significant loss of quality. The Qwen 35B is pretty fast in my testing, but MiniMax M2.7 230B is in some ways faster (way fewer tokens to arrive at an answer) even though it is much larger.
Re:Gemma 4 12B: A unified, encoder-free multimodal model
by SwellJoe on Wednesday June 03, 2026 @10:03PM
(#48392746)
Qwen 3.6 35B-A3B with MTP at 8 bits is blazing fast, something like 50-60 tokens per second. That's plenty fast for interactive use, so I haven't tried lower bits. Unfortunately the MoE is notably dumber than the dense model (for the case I have data about...I've been benchmarking models for security vulnerability scanning, and 27B is notably better on hard bugs).
The dense model is almost usable, but feels really sluggish, even with MTP. I think it's about 12-15 tokens/second on the Strix Halo. Slow enough to where I'd rather pay to use a cloud model.
I might try the 6-bit version of the dense model to see how it behaves, though. Maybe it'll retain its bug hunting abilities while making it fast enough to use interactively and not take all day for benchmark runs.
Re:Gemma 4 12B: A unified, encoder-free multimodal model
by hedgehog on Wednesday June 03, 2026 @11:09PM
(#48393228)
Same chip, with a 6 bit 35B and 8 bit KV cache I see about 500 prefill and 55 decode at 30k into the context window. MiniMax seemed a bit lower token rate but much, much less prone to 40k tokens of monologue before generating an answer. A pattern I like is to use a smaller model to do most execution and then a larger model to review transcripts and output and do any fixups and tooling improvements (this is all batch jobs so all I care about is overall throughput).
Re:Gemma 4 12B: A unified, encoder-free multimodal model
by milch on Thursday June 04, 2026 @01:44AM
(#48394489)
What hardware do you need to run MiniMax M2.7 230B locally?
Even batched usage needs to be fast enough to deliver results in a reasonable time. Overnight runs are useful, 24 hour runs are...less so.
Anyway, most of the time people are talking about interactive use, and there's currently an upper bound on how large a model can be for local hosting on a reasonable budget (i.e. not a crazy amount more expensive than what a high end developer desktop or laptop costs). The sweet spot is probably currently the big Qwen 3.6 or Gemma 4 models, which are in the ~60GB range for 8-bit quantization plus a large context.