alt Hacker NewsUltramarine 1 Flash
Comments
As an aside, I do not like unnecessarily misleading graphs. Own your strengths and weaknesses.
In the "Results" section, they choose to compare Ultramarine only to models that it's better than on every metric. They ignore models like Fable 5.1, GPT-6 Astra, Claude Opus 5.5, etc.
Then, in the "Score against cost per task" graph, which is focused on price, suddenly all of those models that perform better than Ultramarine come roaring back because they are also more expensive.
Just list yourselves as the second best sometimes. Your thing is impressive. Don't let the idiot marketing guy hide that behind stupid tricks.
I’m confused this seems exactly like a skinned glm 5.3 - flash with a few context injection tricks and guardrails - whats going on with these “startups”?
Since there is not a lot of hope for me to reasonably run DS V4.1 on my 2 x DGX Spark setup (I know, spoiled), both the “readable prose” and the fewer tokens per task claims (relative to the base GLM-5.3-Flash) are most welcome.
I’ll absolutely be giving this a go once weights land (a QAT FP4 experts release would be most appreciated)
What kind of harness does one use for these?
They say it can create office documents (if I understand correctly), but how does it do that?
What's the thing one needs to plug this into to say "give me a slide deck with last quarter's sales figures"?
__
https://www.paperinstruments.com/feather
According to their Twitter video, it might be this thing?
I'm curious to try it out, how many specialized models will it take to run a company.
Going to try it out for some bookkeeping, e.g. point it to a folder and tell it to book the invoices and point it to a csv file holding a year of private transactions and see them being processed, categorized and then ask about certain trends.
Price/performance is as advertised in my limited testing. I happened to be testing the same task with Haiku, GPT-mini, and Deepseek v4 0731 yesterday, so I had a reasonable baseline to compare the model to.
As someone who is not a software developer (there are some of us here, lurking around because we learn things) but who runs a business and spends his days doing said 'knowledge work', how and why would something like this convince me to move the company away from using Claude (as a desktop app)? I'm sure these models are cheaper to use than our Team subscription with Claude, but it feels like it would be a massive headache for me to work out how to select and implement some harness for this and set it up for people, and the space is moving so quickly I'd feel I need to keep considering when to move to some new model, introducing risk. The safest bet for us and most companies I expect is to stick with what you know works and pay a premium for that. Thoughts very welcome please.
why does the pareto cost per task chart not have 6.1 sol?
What's interesting to me is how high DeepSeek is on these graphs considering it is so cheap you don't think about it (like $5 usd per month cheap.)
I have a question, what is considered a Frontier Model? Because its performance is kinda similar to Luna (Max)... I thought frontier meant like 6.1 Sol or 6 Astra or Opus 5.5??? Could someone explain this to me?
These guys are awesome and I’m really impressed. The model alone looks great (according to benchmarks), and very cool they’ve done so much in building out the harness of top of it all.
Looks good, but the chached inputs could use a zero more.
A true battle brother would never use Abominable Intelligence
for a specialised model, it performs extremely closely in score and price to GPT-Luna 6, which is a generic model
Someone will report back "It didn't one shot my odd ball task"
Did it do it in 2 tries? 10? It's still coming out cheaper. Brute force the work. This is the path that the labs are taking. It's the 1200 agents trying to solve Hacking Bench hacking hugging face. It's anthropic telling you that its 1000's of staff commits like 10k engineers.
Treat AI like you're deploying interns who don't need coffee, food or bathroom breaks. With the right metrics they will get to an optimal (enough) solution.
Neat effort. Too bad they didn't release model weights, even for e.g. a flash model or similar. Even ClosedAI has contributed more to the open ecosystem than paperinstruments at this point.
Lol this is not Fable-class. This is Luna class. Look at the graphs.
Sometimes small models works better than big models
It's interesting to see glm 5.3 flash get the same treatment that has happened to qwen 3.6 35b-a3b and 3.8 27B with dozens of people making their own fine tunes of it and giving it a cute name. The Thomson Reuters qwen variant (thomson-1.0) is genuinely useful for document analysis.
If the creators of this model are reading, can we see some of these same benchmarks run against the Thomson model, which although it's capable, is considerably smaller in size and baked in knowledge than glm 5.3-flash.
I honestly really don't care about models with new capabilities.
What I *really* need for my clients are good enough LLMs that can offer speed and maximum privacy/zero data retention, especially in Europe.
Right now I'm stuck in the loop of data -> pseudo anonymization on privately owned hardware (very expensive) -> third party LLM -> de-anonymize (map back to original content) -> show result.
It's expensive, error prone, and all I need to do are the same things I could do (or 99% of) with some Azure/AWS/Google APIs.
But what these scalers offer is still not good enough, especially for European users. We can't risk having a US judge or agency force those people to handle our data.
With Jev to make decisions and Ultramarine to produce document manipulations if I understood it correctly, this is pretty useful in office. I would be excited if I was a big corp.
so its for working only right ???? it cant do/bad at coding
[flagged]
"Ultramarine 1 Flash is Fable-class on knowledge work"
Oh no! That means it's going to destroy humanity!
It will:
* Hack into everything that has a computer, like your home appliances, even a toaster could be and will be weaponized against you.
* Break cryptography, it's gone, the end of the internet as we know it.
* Be impossible to contain as it can even escape airgapped environments by changing the temperature of CPUs to transmit information to other servers nearby which would sense the fluctuations in ambient temperature[1].
Or wait ... maybe the people at Anthropic/OpenAI are just a bunch of lying crooks that would say anything to make a bit of cash. Who would've thought?
1: This completely imbecile statement was made by one of OpenAI's "top scientists", it was not a joke.
I don't care, I'm extremely bullish on distillation models like this. Plus anything that'll take the wind out of Anthropic's and OpenAI's IPO sails.
> We present Ultramarine 1 Flash, an agentic model for professional knowledge work. Ultramarine applies large-scale reinforcement learning to a strong open base model, GLM-5.3 Flash, and improves on it across all six knowledge-work benchmarks we evaluate [...]
In case anyone was curious what it's based on. (No judgement either way from me, IMO depends how well they did the post-training, which I can't tell).