Open weight US based Opus 5.5 within 12 months.
Seems unlikely. Maybe open weight at all near there? That'd be great.
Frontier and open weights? Yes.
Frontier and _US_ open weights? No.
I'd expect a US open weights provider to suddenly get bought or go non-open if they get near Opus 5.5 anyway.
(I am assuming Inception is an American company, not sure TBH)
My read on Diffusion LLMs has been that it's cool tech, but only useful for things like like code autocomplete or editing a block of text.
They don't beat autoregressive models on quality per compute, assuming you are serving lots of users and can batch requests.
It's not as good as the current frontier, to be fair. I don't use it to help me think through architecture, for example. But I have found to work well for at least like half of coding tasks once architecture and patterns are in place. Its advantage is that it's very very fast.
but to your point , yeah diffusion models can be kind of brain dead without proper coaxing.
my ideal arrangement is a smarter/bigger model supervising lots of proposals from diffusion models. The diffusion models spit out a big proposal each, the bigger model can vet something as minor as the headline or filename for each proposal, time still ends up being saved.
I'm not saying that any of them are to the level as a deepseek v4.1 flash 1017 or similar but they're catching up. The most recent huge mistral release is also grounds for an optimistic view.
Without hundreds of millions in 9 days a small team is producing a model that can beat DeepSeek 4.1 and a platform to tune it.
Optimistic for the future.
MiniMax M3.1 Flash is 428B total parameters, so the size ratio is more like 1.4x. Model card here: https://huggingface.co/MiniMaxAI/MiniMax-M3.1-Flash
The number HF calculates from the safetensors metadata is often wrong (e.g. counting u8-packed 4-bit weights as half a weight each) so you need to take the number from the actual model card.
So why should I pay for 3-20x more for a few small percentage points in preformance. (I would prefer DS did some things better, but largely I can force my desired behavior), but I could only have access to Deepseek 4.1 and still beat almost everyone using opus/fable/astra on a productivity basis, I need automation tools not "do my entire job so Im worthless tools", and DS has already more than gotten there for me.
No domestic good LLM for me, either.
"I bought this house with my own savings plus a small million dollar loan from mum and dad."
Houses are not built from scratch either. The know-how, power, water, roads, materials…
Is the same true for distillation?
edit: nm I see its a bit big
we've also contracted hundreds of local record store guys with deeply knowledgeable music taste spanning post-punk, classic jazz, early hip-hop, and deep-cut experimental vinyl. This has resulted in the model returning judgmental responses up to 23% more often which we are working on addressing.
I feel like the gap between Opus and Grok is about the same as the gap between Grok and Mistral. I didn't revert its code or anything, but I also didn't refill the initial OpenRouter budget when it ran out.
At the same time, I guess Opus 4 level is pretty good for a Western open weights model. 2027 is going to be amazing.
So the legendary skill of Charles Duchemin is now widely available? What a time to be alive.
cries in GPU racks
*Shrugs*
DeepSeek would never block me
… I can’t wait for custom resizable models!
FWIW: like the Sims where you have certain levels where you move one up and others non-linearly go down e.g. Speed, Intelligence, Conciseness, determinism, trainability