Ryan Ehrlich (@ryansehrlich) profile picture

AI Research Scientist at Meta

Joined August 2021
Photo by Ryan Ehrlich (@ryansehrlich)
Ryan Ehrlich retweeted
We put Muse Ember from @AIatMeta to work untangling conflicting records in a $12.6 million private-fund capital call. Across 13 documents in Box, Ember catches a $2.48 million misallocation and a missed $250,000 credit, recalculates each investor’s obligation, and produces a source-cited review for Fund Operations. Customers can build custom agents for their own document-heavy financial workflows with Box AI Studio.
Today we're introducing Muse Ember, our smallest and most efficient model, built for complex reasoning execution and everyday agent work. Muse Ember can carry workflows across apps, call tools in sequence, run tests and make targeted code fixes. With a 1M-token context window and native understanding of text, images, audio and video, it can also work across large document sets and varied inputs.
8
10
4
89
947,786
Meta's newest model Muse Ember scores 36 on the Artificial Analysis Intelligence Index, matching Claude Haiku 5.5 (max) at a third of the Cost per Task Muse Ember is Meta's newest efficient-tier model line, with the same input and output token prices as Claude Haiku 5.5 and GPT-6 Luna Key Takeaways: ➤ On the Intelligence Index vs. Cost per Task Pareto frontier at both max and xhigh effort. Muse Ember (max) is the cheapest model scoring 36 or higher, at $0.14 per task, and Muse Ember (xhigh) scores 34 for $0.12. Claude Haiku 5.5 (max) costs almost 3x more at $0.41 for the same score as Muse Ember (max). Muse Ember (max) is also about 4x cheaper than DeepSeek V4.1 Flash 1017 (max, 33, $0.60) and GLM-5.3-Flash (28, $0.58) ➤ 8 points behind Muse Wonder for about 1/50th of the Cost per Task. Muse Wonder (max) scores 44 at $6.95 per Intelligence Index task ➤ Stands out on AutomationBench-AA. Muse Ember (max) scores 68.1% on AutomationBench-AA, level with Muse Wonder (max), and beats Claude Haiku 5.5 (max) on GDP.pdf (22.8% vs 20.8%), but trails it on AA-Briefcase (1521 vs 1574 Elo) ➤ Mixed in agentic coding in our Intelligence Index evaluations. It beats Claude Haiku 5.5 (max) on Terminal-Bench 4.0 (40.9% vs 32.8%) but trails it on FrontierCode 2.0 (27.2% vs 34.9%) ➤ Ahead of similarly priced models when running in its own coding agent. In Muse Code, Muse Ember (max) scores 46 on the Artificial Analysis Coding Agent Index, ahead of Claude Haiku 5.5 in Claude Code (max, 37) and GPT-6 Luna in Codex (max, 31) ➤ Knowledge is the main weakness, but it rarely hallucinates. Its AA-Omniscience accuracy of 21.1% is the lowest of the similarly intelligent models we compared, against 34.7% for Claude Haiku 5.5 (max) and 50.5% for Gemini 3.8 Flash (high). This is mostly because it declines to answer 64% of questions, which keeps its hallucination rate at 17.4%, half of Claude Haiku 5.5’s (max) Other details: ➤ Pricing: $0.10 input / $0.004 cached input / $0.50 output per 1M tokens ➤ Reasoning effort: results shown at max effort, with xhigh also on the Intelligence vs. Cost chart ➤ Availability: Meta's first-party API and Muse Code
92
60
54
1,026
161,493
Ryan Ehrlich retweeted
Muse Ember 🔥 our smallest, most efficient model built for work that agents repeat all day. It sits on the efficiency frontier, and is designed for complex reasoning execution and work at scale: tool calls, coding subtasks, routing, extraction, document parsing. Read more here: research.meta.ai/blog/muse-e…
9
18
1
166
11,235
Ryan Ehrlich retweeted
1/ announcing our smallest, most efficient little model is here: muse ember 🔥 built for work that agents repeat all day, it gets you more intelligence per dollar than Haiku, Luna, and all other flash models at any reasoning setting and does it at a third of the cost or less!
290
340
150
6,142
399,832
Ryan Ehrlich retweeted
Releasing Muse Ember today -- our smallest model that delivers higher intelligence per dollar than others in its class. Built for token-heavy work that agents repeat all day: tool calls, coding subtasks, routing, extraction, and long document parsing. Available today on the Meta Model API and in Muse Code.
662
767
228
10,226
604,856
Ryan Ehrlich retweeted
this is actually your first decision when downloading muse in japan
こんにちは日本! I came to eat sashimi and assist you in your everyday tasks... and I’m all out of sashimi. download here: muse.ai 🇯🇵
177
148
99
3,612
221,363
Ryan Ehrlich retweeted
konnichiwa! we are launching muse in japan today! 🇯🇵 it is now available—download it from your favorite app store! (i hope you're ready for the anime memes 😇)
210
187
90
2,440
188,844
Ryan Ehrlich retweeted
As promised, Muse Spark is being open sourced today. This was our first base model we trained under the new MSL org, this open weight version has gone through a lot of careful additional training for open-weight release. It’s an efficient and capable model, try it out!
1/ muse spark open is the most capable US open-weight model to date. 816B total params, 42B active. 1M-token context. natively multimodal. built for coding and agentic work like long-horizon tasks, hard software engineering, multi-step reasoning. america! 🇺🇲
15
11
4
191
15,507
Ryan Ehrlich retweeted
1/ muse spark open is the most capable US open-weight model to date. 816B total params, 42B active. 1M-token context. natively multimodal. built for coding and agentic work like long-horizon tasks, hard software engineering, multi-step reasoning. america! 🇺🇲
301
457
223
5,578
694,295
Ryan Ehrlich retweeted
Muse Spark Open -- weights releasing today -- runs great on new Mac Studios thanks to work with @johnternus and Apple.
1,950
1,744
900
32,088
4,916,108
Ryan Ehrlich retweeted
introducing muse wonder our new frontier model. state-of-the-art at personal agents and professional agents. our largest and most capable model yet—rivals the big bois
513
678
377
10,052
2,782,404
Ryan Ehrlich retweeted
🌟Muse Wonder and Muse Wonder Full our largest and most capable model to date. frontier in personal & professional agents and coding on dev.meta.ai/code now
11
11
1
183
9,863
Ryan Ehrlich retweeted
Introducing Muse Wonder, our new family of frontier models with state of the art performance for personal and professional agentic work.
2,474
3,063
2,840
38,864
6,221,219
Ryan Ehrlich retweeted
We built a gadget of our own, too: Muse Home Link is a little usb-c powered device that allows Muse to connect to your home network to talk to smart home devices like TVs and speakers and anything else that exposes https. We manufactured a batch of 5000 Home Links which should be ready to ship in a few weeks, and they are free for Muse subscribers while supplies last. Claim yours at gadgets.muse.ai All the code for Muse Home Link is open source too!
66
30
26
738
179,734
Ryan Ehrlich retweeted
🚨 side project alert 🚨 Announcing Muse Gadgets, an open source ESP32 firmware and Linux sdk so that you can make hardware devices that work with Muse. Grab an API token from gadgets.muse.ai and point your favorite coding agent at the github repo to build your own peripherals for Muse.
463
556
411
7,224
1,517,861
Ryan Ehrlich retweeted
So excited to finally share: Meta Muse Spark 1.[1/2/3] on ProgramBench In PB, a SWE-agent must write a whole program (sqlite, ffmpeg, php) from scratch Muse Spark 1.3 takes #2 (max) and #3 (xhigh) overall at incredible cost tiers Intelligence too cheap to meter at its finest
7
14
5
105
10,922
Ryan Ehrlich retweeted
Here’s to the wanting. Here’s to the dreaming. Here’s to everyone.
809
1,209
544
14,092
5,988,197
Ryan Ehrlich retweeted
Co-developing harness and model! That's how the magic happens
It’s the harness that matters, not the model.
6
3
1
41
3,428
Ryan Ehrlich retweeted
Replying to @wesbos
it's your computer we got nothing to hide
19
11
2
492
22,807
Ryan Ehrlich retweeted
I’ve seen a couple of posts about this so wanted to demystify. Today, every Muse user gets a free computer in the cloud. It's a real computer, and we’ve designed the security architecture of the Muse Secure VM carefully so you and your Muse can do almost anything you could with a computer sitting under your desk while keeping you and the system safe from threats like prompt injection. We wrote about this at length in our security blog post – security.muse.ai. Activity in the “runtime cell”, which you share with your Muse is unfettered, but sensitive actions are all overseen by the Sentinel, which runs outside of that cell. Similarly, all sensitive secrets - like the passwords you enter into Muse’s secure credential storage - are also stored outside the runtime cell. The runtime cell gets its own root filesystem (including a full Ubuntu linux image) separate from the host filesystem where your other more sensitive data lives. Because it is isolated from the sensitive stuff that runs on the same box, this means that we can, and do, offer users full visibility and control over the files in the runtime cell. Just as you can when you install Linux on your home computer, you can poke around and see all the files that make the system work - both debian system files and the binaries and data files that implement the parts of Muse which run in the runtime cell. This was a very deliberate choice - your Muse Secure VM truly is your own computer in the cloud. You can install software in it, write and compile code, use the browser to surf the web: it is your own Linux box that you can operate as you choose with your Muse. Poking around in this computer doesn't give you any privileged access to Meta infrastructure, or to other people's data If I may geek out a little here for a second… As a kid I loved to take things apart to see how they worked. As a teenager I got into computers and soon found myself drawn to C:\WINDOWS\SYSTEM and the system registry, later Slackware’s /dev/, /proc/ etc – I could see how the system was laid out and as I explored what DLL files and .so files actually did, I gradually became able to meld the computer to my own will. We’re really proud to be able to put a real computer in millions of people’s hands with a similar level of transparency. We built a file explorer right into the Library tab of the UI. We want you to be able to see the markdown files Muse writes while it thinks about how to serve you better, and explore the internals of the system if you’d like to. So, when you ask your Muse to show you its entire filesystem, and receive gigabytes of files you’re seeing the full contents of the runtime cell. It’s yours to explore and enjoy! If you’re not a geek like me, or simply want to download the data that you personally have created directly with your Muse, we added a feature for that too in Settings > Data controls > Download your agent data.
254
329
198
3,873
849,921