Two stealth models just appeared on the leaderboard, both ahead of Fable 5.2
cheptsov
132 points
45 comments
November 19, 2026
Related Discussions
Found 5 related stories in 151.3ms across 13,334 title embeddings via pgvector HNSW
- Saltline: First-person shooter built by Fable 5.2 dr00tb · 56 pts · November 10, 2026 · 51% similar
- GPT-5.6 vs. Claude Fable 5 for Physical AI, which performs best? mbauman · 90 pts · July 29, 2026 · 50% similar
- Fable 5.1 World Modeling surreal_ · 192 pts · September 02, 2026 · 50% similar
- Memdoor: A team of 5 Qwen3.8-27B models beats Fable 5 on LiveCodeBench Hard guregodevo · 162 pts · October 31, 2026 · 50% similar
- New stealth model: Union Alpha jlaneve · 17 pts · September 16, 2026 · 49% similar
Discussion Highlights (18 comments)
_s_a_m_
some obscure design benchmark
skinfaxi
"Random AI lab makes up a new benchmark that puts at least one of its models ahead of Fable 5.2"
cronin101
Saying this when none of the gallery examples show the new model's output is the ultimate "trust me, bro" move.
akutlay
It is interesting that the Kimi model gets a very high “slop score” but “professional critiques” gave it a pretty high score. Do critiques have no taste or the “slop” is not slop anymore?
alex7o
Glm 5.5 I bet, but still too early to tell
barrenko
This Thanksgiving period will be a cambrian explosion of capable models.
irthomasthomas
Deepseek V5 trained on Huawei Ascend?
gorgmah
I don't understand, where are those models available in stealth mode? Usually people do this on openrouter but I couldn't find any model called "gestalt". Does anyone know or is this just black hat marketing for this contralabs platform, and the model does not even exist?
petesergeant
This is a clever bit of content marketing
gajus
To clarify, it is not two models. Its a single model (Gestalt). The other bar on the benchmark is a human created dataset. Suffice to say we didn't expect anyone to beat it, but here we are. And no, it is not Contra model.
curiousgeorge10
Very curious to find out which model that is :O
TheAtomic
That web site is insufferable.
ForHackernews
"on the leaderboard [that we created so we could lead it]"
vonneumannstan
Why would this not be GPT 6.2 Astra?
radzserg
who is --John Galt-- Gestalt?
lyw6
interesting! any model performing better than human created outputs is very unexpected. i wonder how fast these models are gonna get better, thinking about the progress we made in a year on coding, design would be soon too
barbacoa
It is possible that anthropic/open ai may no longer release their best models due to their massive size and cost of running, instead releasing distilled version for public use.
algoth1
The name gestalt indicates it’s probably a router model like paretto by unbiased