Union Alpha
stealth · released Sep 16, 2026
- Type
- Closed
- Input
- None held
- Output
- None held
- Cached
- None held
We don't hold a list price for this model yet · hosted only — we have no record of published weights
Our take
Written Sep 21, 2026Union Alpha is a text-and-image model with no download and no host we can point you to, so there is no route to it from here. Its only measured quality is a set of LiveBench scores, and its licence is not disclosed.
Consider it only if you already have a way to reach it and can judge the output yourself, and only where the maths or language scores are the signal you care about. The maths result is the standout, at 95.29% on competition and olympiad tasks. Skip it if you need to run it yourself or reach it through a host, or if you need to know what the licence permits.
The case for it
- 95.29% on LiveBench Mathematics, which covers competition and olympiad maths rather than everyday arithmetic, so it is the score to weigh if that is your work.
- 85.89% on LiveBench Language, covering comprehension rather than generation, so it is a signal for reading and understanding tasks rather than writing ones.
- 82.15% on LiveBench Coding, covering code generation and completion rather than fixing issues in an existing project, so treat it as a signal for writing new code.
The case against it
- We list no download for it and no host serves it, so there is no route we can point you to.
- The licence is not disclosed, so commercial use and redistribution are unverified in our data.
- Instruction following and agentic coding are the weaker scores: 59.53% on constrained-rewriting tasks and 54.7% on agentic coding run inside an agent harness, so that second figure is a result for the model inside that harness.
How good is it?
EverydayGeneral questions and everyday reasoning
Not yet scored on Arena Text (overall). It is on LiveBench Mathematics, in 12th of 58 with 95.29.
CodingWriting and fixing code on its own
Not yet scored on Arena Coding. It is on LiveBench Coding, in 8th of 58 with 82.15.
AgenticPlanning, calling tools, staying on task
Not yet scored on Arena Agent. It is on LiveBench Agentic Coding, in 25th of 58 with 54.7.
WritingDrafting and rewriting prose
Not yet scored on Arena Creative Writing. It is on LiveBench Language, in 9th of 58 with 85.89.
Boards this model appears on that none of the ratings above are built on.
Every published score for this model8 scoresEvery figure we hold, from 8 boards, with who ran it and a link to the source — including the boards no rating above is built on.
Where to get it
We hold no priced listing for Union Alpha.
There is no copy to download and no host in our price data, so stealth is where to look. We watch OpenRouter, the provider APIs we track and the LiteLLM price set; this version appears in none of them, which is a gap in what we collect rather than a statement about what stealth sells.
When we formed this view
Recent changes
What moved
first indexed by our pipelineEach date is the day we first saw the change, or the day the maker announced it.
What we do not know about this model yet
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- We don't hold a list price for this model yet — the gap is ours, not the lab's.
Licence and identifiers
What the licence allowsWe hold no licence record for this model. Inside are the identifiers you need to pull it — its Hugging Face repo where we have one, our slug and a machine-readable card.
Licence
We hold no licence record for this model, and no record of published weights either — so we can neither summarise its terms nor point you at the weights.
Identifiers
- Takes in, gives back
- Text and images in, text out
- Catalogue slug
- stealth-union-alpha