Ohmygoodness, the results look amazing. I felt so sad for them when their first weights dropped, a lot of other great models came out around then and they didn't get the chance to shine even if for a week or 2. But looking at these results, this is really good stuff and a decent alternative to DSv4.1-F1017. Also the best and most detailed comparison I have seen from a lab across models.
Jackobrien
The best open models today have been “good enough” to do most knowledge work for a few months. This might be the first US one good enough too. That’s a big deal.
Open weight US based Opus 5.5 within 12 months.
show comments
epsteingpt
It's a nice, domestic open weight model.
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.
show comments
itkovian_
The most important thing here is training was entirely nvfp4 - this is best model by far that has ever been trained in that precision and explains why they were able to do it so fast
show comments
a11r
Very refreshing to see benchmarks where the reporting team is not the clear winner. They are claiming to be better than GLM 5.3 flash, which is impressive if it holds up outside of benchmarks.
tcdent
Is it SOTA yet? No. Are the domestic open weights labs demonstrating incredible velocity? Undeniably.
show comments
syntaxing
I can’t seem to find the cost? The craziness of deepseek flash 1017 is how ridiculously cheap it is for how capable it is.
show comments
cauliflower99
As someone who knows nothing about wine except I like drinking it - that wine test is fascinating.
ddxv
This is amazing! Can't wait to trying this out. Also, crazy that I found this on LinkedIn with more upvotes and discussion than the HN thread has.
alexandra_au
Damn their last "Inkling Small" model was only 266B and now this is more than double.
show comments
Neywiny
I feel like this is a dumb question, but why are we spending any effort on wine comparisons? Is there a way to get rid of experts on these to just be ones for code?
show comments
maelito
Looks better than mistral 4 on all benchmarks ?
show comments
rocha_lucas
I couldn't pay attention on the content, i was playing with dots
Tade0
> We even tested Inkling 2 Small against some of our researchers in a wine tasting event. It outperformed them using vision alone, correctly identifying three wines from a single picture in 19 out of 20 trials.
So the legendary skill of Charles Duchemin is now widely available? What a time to be alive.
lostmsu
> Small: 594 billion parameters
cries in GPU racks
show comments
ygouzerh
Interesting, I sent the same wine testing example to Claude Opus 5.5 medium, and it failed, where Inkling 2 Small succeed.
barrenko
They should buy the pleias team and just mog everyone.
AnthonyR
I like the presentation on their website! Super well done
Kuyawa
"Sorry, you have been blocked"
*Shrugs*
DeepSeek would never block me
singpolyma3
So basically comparable to DS4.1? But why bother to make something we already have?
show comments
jazzyjackson
I don’t like it. Poor form to squat on an old companies name like that, one which has a rich legacy in AI. I know the trademark was cancelled so all I can say is I don’t like it.
alfiedotwtf
Looks awesome, but dang a bit too big for my machine.
… 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
Ohmygoodness, the results look amazing. I felt so sad for them when their first weights dropped, a lot of other great models came out around then and they didn't get the chance to shine even if for a week or 2. But looking at these results, this is really good stuff and a decent alternative to DSv4.1-F1017. Also the best and most detailed comparison I have seen from a lab across models.
The best open models today have been “good enough” to do most knowledge work for a few months. This might be the first US one good enough too. That’s a big deal.
Open weight US based Opus 5.5 within 12 months.
It's a nice, domestic open weight model.
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.
The most important thing here is training was entirely nvfp4 - this is best model by far that has ever been trained in that precision and explains why they were able to do it so fast
Very refreshing to see benchmarks where the reporting team is not the clear winner. They are claiming to be better than GLM 5.3 flash, which is impressive if it holds up outside of benchmarks.
Is it SOTA yet? No. Are the domestic open weights labs demonstrating incredible velocity? Undeniably.
I can’t seem to find the cost? The craziness of deepseek flash 1017 is how ridiculously cheap it is for how capable it is.
As someone who knows nothing about wine except I like drinking it - that wine test is fascinating.
This is amazing! Can't wait to trying this out. Also, crazy that I found this on LinkedIn with more upvotes and discussion than the HN thread has.
Damn their last "Inkling Small" model was only 266B and now this is more than double.
I feel like this is a dumb question, but why are we spending any effort on wine comparisons? Is there a way to get rid of experts on these to just be ones for code?
Looks better than mistral 4 on all benchmarks ?
I couldn't pay attention on the content, i was playing with dots
> We even tested Inkling 2 Small against some of our researchers in a wine tasting event. It outperformed them using vision alone, correctly identifying three wines from a single picture in 19 out of 20 trials.
So the legendary skill of Charles Duchemin is now widely available? What a time to be alive.
> Small: 594 billion parameters
cries in GPU racks
Interesting, I sent the same wine testing example to Claude Opus 5.5 medium, and it failed, where Inkling 2 Small succeed.
They should buy the pleias team and just mog everyone.
I like the presentation on their website! Super well done
"Sorry, you have been blocked"
*Shrugs*
DeepSeek would never block me
So basically comparable to DS4.1? But why bother to make something we already have?
I don’t like it. Poor form to squat on an old companies name like that, one which has a rich legacy in AI. I know the trademark was cancelled so all I can say is I don’t like it.
Looks awesome, but dang a bit too big for my machine.
… 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