Meta's Muse appears to use an OpenAI model labeled muse-special

Meta's Muse appears to use an OpenAI model labeled muse-special mouse.dev

While Muse was building my website, one subagent ran on a model called azure/muse-special. So I dug through the filesystem again.

143 points•Aeroi•5 days ago•51 comments•

#muse #special #muse special #appears #labeled #metas #openai #metas muse

51 comments

Aeroi5 days ago

author here. I kept digging through the Muse filesystem after my first post hit the front page here this week.

Going through my logs I found a background agent that used a model called azure/muse-special while building my website. I found this interesting and dug a little deeper.

The transcript and daemon binary point to an OpenAI model running on Azure. Still unclear which one or why it was selected.

The runtime also ships with an Anthropic client and a catalogue listing Claude, GPT and Kimi models. I didn’t observe Claude being used in my own sessions, but this is my Part 2 of exploring the Muse file system.

original post: https://x.com/heypeterjames/status/2103545183400800746

pete at mouse dot dev

haolez5 days ago

Nice work!

Are you sure this isn't just a custom model that is API-compatible with OpenAI?

vineyardmike5 days ago

This feels like the far more likely option.

The Muse public APIs seem to be heavily inspired by OpenAI's, they even support the richer "responses" API. They have a similarly shaped compaction API, they have the same "encrypted" reasoning. Even the Muse harness seems heavily inspired by (open source) Codex. Considering Azure's historic involvement w/ OpenAI, it seems more plausible that Azure is used as spill-over capacity, and the same Azure infra used to encrypt GPT models encrypts Muse models.

My guess that the generally visible Anthropic/OpenAI details are leftovers from the whole meta "move fast" behavior. There are a few rough edges on the product where it leaks internal codenames (eg. signing in w/ WhatsApp required consenting to using "hatch" on one screen), so I'd believe that this was a leftover on the VM from the prototyping stage, before the Muse models were ready, or because the dev's got to benchmark against different models.

Aeroi5 days ago

I'm not no. Someone from meta would have to chime in... but its very openai shaped and served differently from all the other models listed in the daemon, under a mysterious name.

a fine-tuned OpenAI model on Azure for the purpose of compaction or something could make sense I guess but that would still be OpenAI's weights with meta ontop, and they already have a compaction model served under azure/avocado-compaction-v1

embedding-shape5 days ago

If it's using the same weird (and awful) "backend prompt encryption" pattern and mechanism that Codex also uses (https://news.ycombinator.com/item?id=48905028), then I'd also say it points to OpenAI being involved. Hopefully that shit isn't becoming a more popular pattern, absolutely awful for troubleshooting stuff.

binlog5 days ago

Why host it on Azure though?

alexgoodhart5 days ago

What endpoint would Claude be calling?

Tiberium5 days ago

I don't really see evidence that it's an "OpenAI" model, the title is IMO very misleading.

Aeroi5 days ago

Meta already serves its own models on Azure under their real names. azure/avocado-compaction-v1 and azure/avocado-memory-flush-v1 are in the same catalog. Those sessions do not come back as gpt_responses_v1 items with rs_ ids and encrypted reasoning.

VygmraMGVl5 days ago

Avocado and Muse are different models. Muse uses the openAI API.

https://dev.meta.ai/docs/overview

Source: I work on AI at Meta.

catchnear43215 days ago

router makes sense. fill a gap with someone else’s model wearing a mask, swap it out with your model not wearing a mask when (or if) you reach that.

definitely tracks with alexandr’s hand-wavey influencer turn.

6thbit5 days ago

It'd be hilarious that meta chose to pay openai. If there was such a deal, would it come to light in any public/official filings?

binlog5 days ago

Depends on the size of the deal. OpenAI would have to disclose it in their S-1 if it crossed the threshold of being material information for investors.

enjoyyourlife5 days ago

Isn't S-1 just for IPO? Wouldn't it be in the 10-Q?

micromacrofoot5 days ago

meta already pays openai and anthropic for employee use despite having their own models available

meric_5 days ago

No reason for it to. Software companies use products from other software companeis all the time.

manav5 days ago

Muse 1.3 spark occasionally spits out Chinese character responses to me like internal instructions. “Go fast” or “Get help”…

sroussey5 days ago

I’ve had Claude opus do that too.

wincy5 days ago

I had a totally benign chat with OpenAI and it titled it as “amateur porn” in Chinese characters, it was very alarmed when I pointed the conversation name out to it. It almost never misses these days, but when it does the failure modes are very strange.

georgemcbay5 days ago

Not sure what model(s) they use but this recently happened in a TV interview with an "AI actress".

https://www.reddit.com/r/oddlyterrifying/comments/1wljyvf/al...

tclancy5 days ago

Got Russian from it once. As a child of the '80s I jumped right on it, but the explanation seemed believable.

Still sleeping with one eye open.

ChickeNES5 days ago

A Hindi word in the middle of a normal reply for me (but once I translated the word it was right in context heh)

Read the full thread on Hacker News →

51 comments

Related stories