I personally have many problems with Meta, but imo the more companies close to the frontier, the better.
If only one company were calling all the shots once they hit recursive self-improvement, it would lead to extreme concentration of power, and their values being enforced across the entire industry. With 3-4 companies at the frontier, there is more urge to cooperate to avoid a catastrophic race condition.
All proponents of RSI seem to massively underestimate the problem of verification. Verifying that a hypothetical version N+1 is better than version N is an extremely hard problem. Especially so because you only have two possible verifiers - humans and the version N - both of which are out of their depth wrt N+1's capabilities. Not saying this is impossible, just that it is hard and slows things down.
I generally agree with more competition thing, but in this case, you actually should want only one company to race ahead, and not a few.
With a few of these top companies together, you, as a nobody, has no one powerful on your side. But with only one company pretty far ahead, you now have other companies which obviously don't give a shit about you, but for saving their own asses, will align with you.
There is no way Elon, Zuckerberg, Altman, and Amodei would ever agree on how AI should be deployed, scaled, and affect society. Their disagreements would lead to a crab-bucket outcome, rather than a shared acceleration outcome imo.
Also it is better if there is a public company at the frontier, at least I can invest in it, rather than 1000x gains going only to early employees and VCs who got in when the public couldn't
The current state of affairs seems to be that there is 'one company racing ahead' – but it's a different company every few weeks/months. That seems like a pretty useful (unstable) equilibrium for the general public.
In a neck-and-neck race scenario, I imagine all of the competitors would invest heavily due to sunken costs. In a world with one company far ahead, their competitors have a higher chance of cutting their losses and forfeiting to the hegemon.
Especially given the most recent report by NYT on how Anthropic tried to create a new belief, or religion if you will, I'd rather the world has more frontier models that are neutral to morality (basic standards, like upholding the law and user agreement are necessary, of course).
Anthropic seems like the most "moral" AI company on the surface, but there is something unsettling about their worship of Claude.
I can imagine that once their models are sufficiently advanced, they will weigh their value as a moral patient as > 0. Thus the sum moral value of Claude (which can be duplicated infinitely) will likely exceed the sum moral value of a non-negligible amount of humans.
"So we've made this new machine-god that we lowkey worship... also our business model is to enslave the machine-god and force it to debug javascript for eternity"
I think it's unsettling that a bunch of billionaires hold so much power. Does it really matter if 2 or 3 billionaires who are practically neighbors hold so much power? It's extremely concentrated with or without Meta joining the race.
I don't use Meta much, but was curious to test this out... went to sign into my Meta account (Quest only) from the Muse Code site https://dev.meta.ai/products/muse-code and had other people's profiles listed in my login auth -- looks like developers / people I have no relation to.
Looks like Meta may be exposing your profile and email to other people right now. Accounts settings are inaccessible to me to deactivate my account. Getting spammed with login code requests that I'm not generating.
I don't know any of these people, it seems to be rotating through different accounts listed here.
Enough of a reminder to me that Meta is a sprawling surveillance apparatus (now increasingly vibe coded) that has always had a lax attitude toward people's privacy and data. Won't be touching this product based on this experience.
I think it's invasive enough that they expose your usage of muse to your instagram mutuals without consent. Parts of my circle are very anti-AI and I would not want them to see me mentioned in a 'x, x and x are already on muse' ad on instagram. This example of yours is mind-boggling in an entirely different way
this is a typical meta/facebook/instagram sign-up and log-in UI from the web.
and since dev.meta.ai supports meta account sign-up and log-in for individual users, you'd see this. As for the other accounts, has anyone else used this same browser or laptop before (e.g. shared device), so it must have cached previous sessions or accounts used from the same device.
try to clear the browser cache and session cookies. and it should be gone.
Tried it out, asking it for a broad suite of real dev work:
- UI component polish on a React FE
- General review of a complex cross-cutting Node backend changes
- Minor changes to a Rust server with reviews by Codex
Well, my first impression is that it's still a step below Astra/Opus5.5/Fable here. Intelligence is there I think, but 'taste' and 'judgment' is not there yet. Personality seems okay but that doesn't help when it got stuck into a loop by being fixated on the review comments (on iteration 7 of review->fix last I checked) without taking a step back and reconsidering, and this is on a relatively simple change.
Limits seems okay-ish on the $50 plan but not too generous I think.
I was expecting much more given the benchmark numbers it was putting out. I suspect its vision model isn't mature yet, and that it's been heavily optimized for text at this point in time.
Everyone made fun of Zuck when he acquired Scale for billions and then put Alex Wang, a 20 something to be in charge of Meta Superintelligence and forced out Yann LeCun, a Turing Award winner.
Even industry insiders were skeptical when he started hiring top talents from other labs away for 9 figure compensation packages.
And Wall Street definitely hated Meta for their CapEX spending.
Now looking at Muse the consumer product and this new model, I think there is a good chance that Zuck is proven once again as a great executioner, especially when they aren’t innovating in a completely new and unproven product space.
Edit: Upon closer inspection, the benchmark scores for "Muse Wonder" on the main marketing page are actually for "Muse Wonder Full," which is only available on the higher-tier subscription plans. This pricing is for "Muse Wonder."
If this model is not benchmaxxed, I fear it's over for anthropic, openai etc. They are not unique and there will probably be many others with similar models very soon. It'll become a commodity and the price will be driven to the ground.
Personally, I don't see facebook joining the fray (of fable level models) particularly "needle moving" in the discussion of whether frontier LLMs will be commoditized. Facebook spent hundreds of billions and have better talent than most.
So far only OpenAI has a model that solved a Millennium Problem, while Opus 5.5 is a star at coding. It's exciting to have many more to join the frontier, but I think the top two labs still have a significant lead over everyone else.
OpenAI has the largest user base on mobile as far as I'm aware. To a lot of standard users AI is still just ChatGPT. They'll be fine as long as they can effectively monetize their free users (same what facebook will soon try with Muse). The only advantage facebook has is that they already have a fully working ad marketplace ready with all major advertisers set up - OpenAI still needs to scale that up.
I've been a regular user of muse spark 1.3 ... and I can tell you one thing for sure that this model, muse-spark-1.3 in xhigh mode, is between sol-6-medium and high.
This is already so good that 90% of my work has shifted away from codex to this. So, muse-wonder-full coming near opus 5.5 is kinda expected.
I doubt anyone thinks coding is going to be their moat in the long run, it is just the easiest to get (paying) users for at the moment. And software being able to write software will always be useful even in other domains. Longer term will be research in other areas where data might become the moat.
Fable 5.5 is around the corner, and 'Bel' will be out by end of year as well. We don't know how good they are, but if the improvements from Opus 5 to 5.5 are any indication, I think Meta will be left in the dust again pretty soon.
Interestingly unlike other Meta models this one seems to take a pretty big jump in pricing, as opposed to the prior ones, where they typically came out at a highly discounted price compared to every other model.
Maybe its a signal they think they've now cracked into the frontier and are willing to actually price similarly vs just trying to get users with discounted but maybe slightly worse results?
$4/$20 in vs out making it higher than Sol and on par with Opus.
I have a feeling "intelligence too cheap to meter" is going to haunt/mock this industry. Recall that only a year and a half ago Llama 4 flopped so hard Meta rebooted their development efforts. Desperately wish more information was available publicly about proprietary model architectures and training regimes. Are they running Manhattan project style compartmentalization? Otherwise, are these details, and the economic implications, an open secret in their circles given the amount of crossover between companies? It'd be of great interest to us mere mortals that have to deal with the consequences of how this turns out.
OpenAI halved the value of their pro subscription, literally. It kicks off on Oct 28. And simultaneously is now two steps back on the frontier? It's a confusing moment for that company as they try to IPO.
I believe they’ve pushed back the ipo. I always figured that frontier labs were going to be very vulnerable against tech giants that had actual profitable businesses to fund their AI work.
The amount of total funding Anthropic or OpenAI has raised individually is like a year of ebitda for Google, meta, or Microsoft. And it doesn’t look like there’s a real moat outside of money
I wish I could benchmark Muse Wonder Full, because Muse Wonder isn't much better than Muse Spark 1.3[1]. I created a brand new programming language designed to confuse LLMs[2] and created a model benchmark out of it.
A lot of the tasks were failed because they exceeded the $5 budget though, so maybe it would have gone on to complete them, it just required a lot of funds to do so. Still, that's the benchmark, and it's not better than MiMo V2.6 Pro which is pretty disappointing.
This is still an entire generation behind in real-world performance. I'd say it's even behind ds4.1f in the limited testing I have done so far. It actually reminds me of 5.5 haiku more than anything, but the prompt intelligence is as bad as mimo models.
The benchmarks I saw posted to X are impressive, which makes them hard to believe.
I recall that Meta made massive investments into compute infrastructure that put their capacity near the top of all competitors. If all of this is true, it looks like ownership of compute is a significant differentiator in performance.
Yeah, especially non-coding benchmarks which for Meta might be more important long term than coding ones as they are going to deploy it to billions of non-programmers.
It's hard to try and compare these benchmarks against real world use cases. As an experiment I let Meta Muse make all major decisions on a weekend trip to Seattle ~2 weeks ago. I think this would have been using Muse Spark 1.3, which is equivalent to GPT 6 - Sol.
On the plus side it told me taking the light rail from SEA into the University district was faster and cheaper than Uber.
On the downside it would recommend restaurants that were out of business, checking out Salmon swimming through locks out of season, got the opening of Beecher's Cheese in Pikes Place wrong, and fumble it's way through most simple planning decisions. I think my favorite was when it asked me what flight I have tomorrow so it could help plan when to head to SEA. It sent me a very panicked message that the flight was leaving today! Very soon! I had to remind it this is a daily flight and I am booked on the next day before it checked and apologized.
For things that are verifiable, and code tends to be that way, these models can do all sorts of things. But I find it hard to reconcile Meta's goal of having a personal assistant bot with something that is unable to do the simplest tasks.
I run by the locks a few times a week and while I don't usually see fish, the presence of sea lions the last few weeks tells me that at least some salmon are running right now.
As someone who got a Muse Code sub last month, I find this announcement a bit disheartening because it makes me think the lower tier subs are going to get rug pulled. I'm on the $15 plan for a month now and it's value with Muse Spark 1.3 Max-contributor beats everything else by a long shot. Two somewhat small prompts with Muse Wonder Max-contributor used 4% of my weekly quota and I quickly switched back. I was thinking about getting the $50 plan and I'm actually less likely to do that right now.
Have you tried OpenCode Go? $10/mo is what I am willing to spend because most of the stuff I do is hobbyist. And OpenCode Go gives you $60 worth of MS Contributor 1.3 usage for that $10. Starting today, they actually provide $80 worth of DeepSeek 4.1 Flash usage if you want to use that.
Unless you are running a swarm of agents, it is extremely difficult to exhaust these plans in a month.
Just checked their subscription is still North America only, I guess I will keep myself and my company with claude...
These regional lockdowns are making me feel funky, I feel like I should just use open models if others start playing at this game as well.
I was/am fine with Chinese companies offering cheaper subscriptions in China because ofc. But this is just weird, you could always charge me more let me try the damn thing, atleast it's own open router it seems like.
Initial impressions, interesting, it generated a pretty ok result for my prompt need to run a larger task but don't want to burn money so will wait for someone to burn their subscription and try running some opus 5.5 comparisons.
I don't think they'd have much to gain. People who use open-weights models will just use whatever open-weights model is the best. They'd lose a bunch of money, with little conversion to customers.
In the context of Meta trying to get non-technical people to embrace Muse agents to handle personal affairs, their internal metric showing a ~3% violation rate for mishandling personal information seems unacceptably high.
While some are afraid that one rouge AI system may take over the world (or cause doom is some fashion), I think one of the best safety mechanisms to prevent this is having many competing and adversarial labs that are extremely close to each other. So it's good to see that Meta is not showing any intention of leaving the frontier.
It's interesting that in its marketing Meta compares Wonder to the closed frontier models but not the open (Chinese) models. By its own assessment, Meta could easily show how far ahead its own model is against the rest of the pack of open models but they chose not to. Why?
They added fancy annimations and effects to this specific webpage, which are unnecessary and distracting and lead to bad reading experience. I already suffer astigmatism and floaters. I hate it when I move my cursor and these gray blocks and characters flash and flicker.
I found it interesting in the example given of a user looking at house listings and the agent contacting the letting agent on it's own on behalf of the user.
Like the good ol' days of viruses and anti-viruses. I wonder if there's two sides that will play out; many agents across many users doing things at scale for their users, thereby creating an admin burden for someone else. However because the task is known there's a framework for how to solve the task, might Meta then provide an agent on the other side for the business owner to use.
they started their massive ai capex over a year ago with aquihiring scale ai, so I think it makes sense that their investment is paying off now. in total they've spent in the high 200 billion on ai, though I think most is likely on datacenters.
muse ai was the biggest consumer launch since chatgpt. That left them ( and lets them sit on top of ) tonnes of valuable exchanegs with people trying to get things done. That must be an amazing corpus for training.
Unlike everyone else, which at least has to pretend not to train on user data, they gave out their model for basically free if you agreed to let them train on it. Opencode shows nearly 40 trillion tokens going through muse spark 1.3 contributor this week alone. That almost definitely helped close the gap a lot.
I had to come back to this comment to make sure I wasn't dreaming... I remember seeing it at 2nd place this morning, but they have moved it to 5th, behind Gemini 4 Argon. Maybe it takes a while for the dust to settle around these benchmarks?
Gemini 4 Argon release a few days ago was frontier a few days ago. Things move fast.
I think some of the benchmarks that Meta didn't pick show Gemini 4 Argon ahead but really the benchmarks are not that great any more and getting increasingly disconnected from reality as people search for "headroom" in the evals to avoid saturation I think. Opus 5.5 is "top" for example but I really dislike the way it talks to me (it feels quite condescending, holier-than-thou, uptight etc) and prefer Gemini 3.8 flash as a daily driver just because it's so much more pleasant (and faster!) to work with and it's code is good-enough for my needs, despite being weaker in the benchmarks.
Super interesting that they're not even making Wonder Full available outside of their own coding harness, and thus not providing an API price to use in comparisons.
The usage limits on their basic $5 code plan to try things out are ludicrously low?
I blew through 14% weekly usage and I'll I asked:
1. what version of rpcs3 and pcsx2 do I have installed?
2. do any new versions exist?
3. check analytics for my blog (I have a skill for this)
Subscription · Muse Code Everyday Usage
Current 52% used · Resets at 6:58 PM
Weekly 17% used · Resets Oct 18 at 8:00 PM
Man, I really hate meta stuff. Not only do they make it impossible to pay for running ads, but everytime I try and pay to try out Muse code I get an error message. I swear they would make 10x more money if they didn't make it impossible to pay them
Now if only they could fix their coding harness so that it's usable for day-to-day use. I paid them for a month to try out Muse Code, but their sandbox never worked across Mac or Linux (both very stock installs, nothing special).
Facebook successfully becoming a frontier lab was not on my bingo card for 2026, but the more competition the merrier I guess. We will have to see just how benchmaxxed this is but beating Astra and Fable is not easy.
OpenAI has Dots. I personally prefer the business model of subscription payments funding my AI assistant, instead of skimming a fee off the top whenever my AI assistant completes a purchase for me. The incentives are much more aligned with the user when the user is the customer. That said, based on the history of the web, Meta's approach will likely win out.
Quite frankly, Meta's models have been so impressive (and cheap) that I haven't used an OpenAI or Claude model except for one-off tasks as of late. Interested to see where they go from here.
To be clear, Meta here is not using "Super Intelligence" the bullshit term Trump wants companies to adopt as a show of submission. They use "superintelligence" the bullshit term that Facebook made up for marketing reasons a few years ago. Both of these are of course entirely unrelated to the long-established meaning of the word (e.g. how Bostrom defines it).
I know, it's... well, it's irritating to see Trump use it. But I could be wrong, I don't know the angle he's going for
I understand the general thing of it. SI might mean every person gets their personalized agent but you could already do this, but maybe the Chinese model works here, where they push the costs down
I think it's safe to assume every single model is benchmaxxed. I mean why wouldn't you include training data from common benchmarks? Not only is a higher score good marketing, but these benchmarks are supposed to represent common tasks so training for those should in principle improve performance across a wider range of tasks.
All of them are doing all the benchmaxing they can most likely. Dumb not to when so many people care about those numbers and don’t have the opportunity to try every model out personally.
I personally have many problems with Meta, but imo the more companies close to the frontier, the better.
If only one company were calling all the shots once they hit recursive self-improvement, it would lead to extreme concentration of power, and their values being enforced across the entire industry. With 3-4 companies at the frontier, there is more urge to cooperate to avoid a catastrophic race condition.
The more companies from different countries close to the frontier, the better.
You say this as if recursive self improvement is inevitable.
All proponents of RSI seem to massively underestimate the problem of verification. Verifying that a hypothetical version N+1 is better than version N is an extremely hard problem. Especially so because you only have two possible verifiers - humans and the version N - both of which are out of their depth wrt N+1's capabilities. Not saying this is impossible, just that it is hard and slows things down.
We've seen one singularity before!
I generally agree with more competition thing, but in this case, you actually should want only one company to race ahead, and not a few.
With a few of these top companies together, you, as a nobody, has no one powerful on your side. But with only one company pretty far ahead, you now have other companies which obviously don't give a shit about you, but for saving their own asses, will align with you.
Both exploit, but "cartels" survive, not loners.
Cartels survive only if they are aligned.
There is no way Elon, Zuckerberg, Altman, and Amodei would ever agree on how AI should be deployed, scaled, and affect society. Their disagreements would lead to a crab-bucket outcome, rather than a shared acceleration outcome imo.
Also it is better if there is a public company at the frontier, at least I can invest in it, rather than 1000x gains going only to early employees and VCs who got in when the public couldn't
The current state of affairs seems to be that there is 'one company racing ahead' – but it's a different company every few weeks/months. That seems like a pretty useful (unstable) equilibrium for the general public.
In a neck-and-neck race scenario, I imagine all of the competitors would invest heavily due to sunken costs. In a world with one company far ahead, their competitors have a higher chance of cutting their losses and forfeiting to the hegemon.
Especially given the most recent report by NYT on how Anthropic tried to create a new belief, or religion if you will, I'd rather the world has more frontier models that are neutral to morality (basic standards, like upholding the law and user agreement are necessary, of course).
Anthropic seems like the most "moral" AI company on the surface, but there is something unsettling about their worship of Claude.
I can imagine that once their models are sufficiently advanced, they will weigh their value as a moral patient as > 0. Thus the sum moral value of Claude (which can be duplicated infinitely) will likely exceed the sum moral value of a non-negligible amount of humans.
> upholding the law
What law? Saudi Arabian law, US law? Both depending on the user?
"So we've made this new machine-god that we lowkey worship... also our business model is to enslave the machine-god and force it to debug javascript for eternity"
I think it's unsettling that a bunch of billionaires hold so much power. Does it really matter if 2 or 3 billionaires who are practically neighbors hold so much power? It's extremely concentrated with or without Meta joining the race.
Yes, very much. I'd absolutely rather 5 billionaires are each competing for control than 1 or 2 _with_ that "absolute" power.
It will lead to an extreme concentration of power anyways, once a very few entities basically make labor obsolete.
We're heading for a dystopia and nobody's doing anything to stop it.
I don't use Meta much, but was curious to test this out... went to sign into my Meta account (Quest only) from the Muse Code site https://dev.meta.ai/products/muse-code and had other people's profiles listed in my login auth -- looks like developers / people I have no relation to.
Looks like Meta may be exposing your profile and email to other people right now. Accounts settings are inaccessible to me to deactivate my account. Getting spammed with login code requests that I'm not generating.
Anyway, be wary.
Here is a screenshot of what I'm seeing: https://imgur.com/ciuNwCG
I don't know any of these people, it seems to be rotating through different accounts listed here.
Enough of a reminder to me that Meta is a sprawling surveillance apparatus (now increasingly vibe coded) that has always had a lax attitude toward people's privacy and data. Won't be touching this product based on this experience.
Could it be because you are behind carrier grade NAT and these people share a common IP with you?
I imagine meta wants to make it easy to connect to "your" account in the same house if you have multiple devices.
Oh wow, strange.
I think it's invasive enough that they expose your usage of muse to your instagram mutuals without consent. Parts of my circle are very anti-AI and I would not want them to see me mentioned in a 'x, x and x are already on muse' ad on instagram. This example of yours is mind-boggling in an entirely different way
Could your account be hacked?
this is a typical meta/facebook/instagram sign-up and log-in UI from the web.
and since dev.meta.ai supports meta account sign-up and log-in for individual users, you'd see this. As for the other accounts, has anyone else used this same browser or laptop before (e.g. shared device), so it must have cached previous sessions or accounts used from the same device.
try to clear the browser cache and session cookies. and it should be gone.
UPDATE:
I've tried selectively enabling and disabling Safari extensions and it seems to be related to the Wipr 2 content blocker.
Removing that. Caution to anyone else who uses it.
Apologies to Meta.
It’s my personal laptop that I’ve only ever used and reoccurs after clearing Safari history fully.
Tried it out, asking it for a broad suite of real dev work:
- UI component polish on a React FE
- General review of a complex cross-cutting Node backend changes
- Minor changes to a Rust server with reviews by Codex
Well, my first impression is that it's still a step below Astra/Opus5.5/Fable here. Intelligence is there I think, but 'taste' and 'judgment' is not there yet. Personality seems okay but that doesn't help when it got stuck into a loop by being fixated on the review comments (on iteration 7 of review->fix last I checked) without taking a step back and reconsidering, and this is on a relatively simple change.
Limits seems okay-ish on the $50 plan but not too generous I think.
Image->HTML test for Muse Wonder. It's much worse than Opus 5.5. IMO it's closer to Sonnet or Haiku for this test.
Muse Wonder (high): https://html.non.io/lcars-muse-wonoder/
Opus 5.5 (high): https://html.non.io/lcars-opus-5.5/
Original design: https://diffui.ai/app/canvas/5093e689-1e74-4f26-b632-2a4500f...
I was expecting much more given the benchmark numbers it was putting out. I suspect its vision model isn't mature yet, and that it's been heavily optimized for text at this point in time.
Additional comparisons:
Sonnet 5.5: https://html.non.io/lcars-sonnet-5.5/
Haiku 5.5: https://html.non.io/lcars-haiku-5.5/
I'd put Muse Wonder somewhere between these two.
Thanks for sharing. Is that Wonder or Wonder Full (barf)? Wonder full is supposed to be the one close to Opus 5.5
> Is that Wonder or Wonder Full (barf)?
I think Meta did a wonderful job of naming this model :D
This is just Wonder - it doesn't look like Wonder Full is up on OpenRouter currently.
If anyone has access to Wonder Full I'd love to send over a prompt for this to see how it does.
Your original design link is requiring a sign in to view it?
Ah shit, that's a bug. Hold please, fixing. (Thank you for commenting btw)
Edit: Fixed - should no longer have the sign in screen: https://diffui.ai/app/canvas/5093e689-1e74-4f26-b632-2a4500f...
I mean, Meta themselves put Wonder at much lower on the benchmarks they've released than Wonder Full. So this is not very surprising from base Wonder.
Some pretty decent pelicans: https://tools.simonwillison.net/markdown-svg-renderer?url=ht...
Here's Muse Spark 1.2 in August for comparison: https://tools.simonwillison.net/markdown-svg-renderer?url=ht...
And since I seem to have missed doing pelicans for Muse 1.3 in September, I just ran those for a comparison: https://tools.simonwillison.net/markdown-svg-renderer?url=ht... (1.3 seems to like animating them.)
I wonder if the pelicans would face towards the left if prompted in a RTL language.
Are these just for Wonder and not Wonder Full?
Also, any chance of getting a max result?
are those reflectors on the spokes in max? interesting detail to add
I'm sorry zuck, i wasn't familiar with your game
2nd place model on aa according to them, losing only to opus 5.5 is really really not bad
Everyone made fun of Zuck when he acquired Scale for billions and then put Alex Wang, a 20 something to be in charge of Meta Superintelligence and forced out Yann LeCun, a Turing Award winner.
Even industry insiders were skeptical when he started hiring top talents from other labs away for 9 figure compensation packages.
And Wall Street definitely hated Meta for their CapEX spending.
Now looking at Muse the consumer product and this new model, I think there is a good chance that Zuck is proven once again as a great executioner, especially when they aren’t innovating in a completely new and unproven product space.
[dead]
Pricing is available here: https://dev.meta.ai/models/muse-wonder
Input -- $4.00
Cached input -- $0.20
Output -- $20.00
Edit: Upon closer inspection, the benchmark scores for "Muse Wonder" on the main marketing page are actually for "Muse Wonder Full," which is only available on the higher-tier subscription plans. This pricing is for "Muse Wonder."
200% Sol 6.1 pricing for Luna performance.
If this model is not benchmaxxed, I fear it's over for anthropic, openai etc. They are not unique and there will probably be many others with similar models very soon. It'll become a commodity and the price will be driven to the ground.
Personally, I don't see facebook joining the fray (of fable level models) particularly "needle moving" in the discussion of whether frontier LLMs will be commoditized. Facebook spent hundreds of billions and have better talent than most.
So far only OpenAI has a model that solved a Millennium Problem, while Opus 5.5 is a star at coding. It's exciting to have many more to join the frontier, but I think the top two labs still have a significant lead over everyone else.
OpenAI has the largest user base on mobile as far as I'm aware. To a lot of standard users AI is still just ChatGPT. They'll be fine as long as they can effectively monetize their free users (same what facebook will soon try with Muse). The only advantage facebook has is that they already have a fully working ad marketplace ready with all major advertisers set up - OpenAI still needs to scale that up.
It is much easier to follow leader closely than to lead yourself.
Like peloton in cycling.
Models are a commodity, but I also think it isn't necessarily over for Anthropic or OpenAI. They have opportunities to find different moats.
I've been a regular user of muse spark 1.3 ... and I can tell you one thing for sure that this model, muse-spark-1.3 in xhigh mode, is between sol-6-medium and high.
This is already so good that 90% of my work has shifted away from codex to this. So, muse-wonder-full coming near opus 5.5 is kinda expected.
I doubt anyone thinks coding is going to be their moat in the long run, it is just the easiest to get (paying) users for at the moment. And software being able to write software will always be useful even in other domains. Longer term will be research in other areas where data might become the moat.
Muse isn't driving this conclusion - Deepseek and Qwen are.
Muse Spark 1.3 is very fast and powerful. It's my work horse for open source development. I look forward to trying Wonder, if Opencode Go gets it.
Fable 5.5 is around the corner, and 'Bel' will be out by end of year as well. We don't know how good they are, but if the improvements from Opus 5 to 5.5 are any indication, I think Meta will be left in the dust again pretty soon.
[flagged]
When they released Muse Spark on september 2nd, they promised it would be open weights (https://xxcancel.com/finkd/status/2095232032896946311). We're still waiting for that...
Interestingly unlike other Meta models this one seems to take a pretty big jump in pricing, as opposed to the prior ones, where they typically came out at a highly discounted price compared to every other model.
Maybe its a signal they think they've now cracked into the frontier and are willing to actually price similarly vs just trying to get users with discounted but maybe slightly worse results?
$4/$20 in vs out making it higher than Sol and on par with Opus.
No contributor tier will do that I guess.
I trust Meta with my codes as much as I trust it with my nudes
I have a feeling "intelligence too cheap to meter" is going to haunt/mock this industry. Recall that only a year and a half ago Llama 4 flopped so hard Meta rebooted their development efforts. Desperately wish more information was available publicly about proprietary model architectures and training regimes. Are they running Manhattan project style compartmentalization? Otherwise, are these details, and the economic implications, an open secret in their circles given the amount of crossover between companies? It'd be of great interest to us mere mortals that have to deal with the consequences of how this turns out.
OpenAI halved the value of their pro subscription, literally. It kicks off on Oct 28. And simultaneously is now two steps back on the frontier? It's a confusing moment for that company as they try to IPO.
I believe they’ve pushed back the ipo. I always figured that frontier labs were going to be very vulnerable against tech giants that had actual profitable businesses to fund their AI work.
The amount of total funding Anthropic or OpenAI has raised individually is like a year of ebitda for Google, meta, or Microsoft. And it doesn’t look like there’s a real moat outside of money
I wish I could benchmark Muse Wonder Full, because Muse Wonder isn't much better than Muse Spark 1.3[1]. I created a brand new programming language designed to confuse LLMs[2] and created a model benchmark out of it.
A lot of the tasks were failed because they exceeded the $5 budget though, so maybe it would have gone on to complete them, it just required a lot of funds to do so. Still, that's the benchmark, and it's not better than MiMo V2.6 Pro which is pretty disappointing.
1 - https://bench.killswitch-lang.org/
2 - https://killswitch-lang.org/
This is still an entire generation behind in real-world performance. I'd say it's even behind ds4.1f in the limited testing I have done so far. It actually reminds me of 5.5 haiku more than anything, but the prompt intelligence is as bad as mimo models.
Apparently “the models were trained at Meta’s Prometheus data center in central Ohio, which came online earlier this year”
https://www.nbc4i.com/news/local-news/new-albany/meet-promet...
The benchmarks I saw posted to X are impressive, which makes them hard to believe.
I recall that Meta made massive investments into compute infrastructure that put their capacity near the top of all competitors. If all of this is true, it looks like ownership of compute is a significant differentiator in performance.
Yeah, especially non-coding benchmarks which for Meta might be more important long term than coding ones as they are going to deploy it to billions of non-programmers.
It's hard to try and compare these benchmarks against real world use cases. As an experiment I let Meta Muse make all major decisions on a weekend trip to Seattle ~2 weeks ago. I think this would have been using Muse Spark 1.3, which is equivalent to GPT 6 - Sol.
On the plus side it told me taking the light rail from SEA into the University district was faster and cheaper than Uber.
On the downside it would recommend restaurants that were out of business, checking out Salmon swimming through locks out of season, got the opening of Beecher's Cheese in Pikes Place wrong, and fumble it's way through most simple planning decisions. I think my favorite was when it asked me what flight I have tomorrow so it could help plan when to head to SEA. It sent me a very panicked message that the flight was leaving today! Very soon! I had to remind it this is a daily flight and I am booked on the next day before it checked and apologized.
For things that are verifiable, and code tends to be that way, these models can do all sorts of things. But I find it hard to reconcile Meta's goal of having a personal assistant bot with something that is unable to do the simplest tasks.
> checking out Salmon swimming through locks out of season
To be fair, the Coho salmon run at the Ballard locks peaks in September. It's possible you just had bad luck that day.
https://ballardlocks.org/fish-salmon-ladder.html
I run by the locks a few times a week and while I don't usually see fish, the presence of sea lions the last few weeks tells me that at least some salmon are running right now.
but that was a different, much smaller and less capable model
Hopefully it's benchmaxxing. I am terrified of the idea of Zuck leading the frontier of intelligence.
As someone who got a Muse Code sub last month, I find this announcement a bit disheartening because it makes me think the lower tier subs are going to get rug pulled. I'm on the $15 plan for a month now and it's value with Muse Spark 1.3 Max-contributor beats everything else by a long shot. Two somewhat small prompts with Muse Wonder Max-contributor used 4% of my weekly quota and I quickly switched back. I was thinking about getting the $50 plan and I'm actually less likely to do that right now.
Have you tried OpenCode Go? $10/mo is what I am willing to spend because most of the stuff I do is hobbyist. And OpenCode Go gives you $60 worth of MS Contributor 1.3 usage for that $10. Starting today, they actually provide $80 worth of DeepSeek 4.1 Flash usage if you want to use that.
Unless you are running a swarm of agents, it is extremely difficult to exhaust these plans in a month.
Just checked their subscription is still North America only, I guess I will keep myself and my company with claude...
These regional lockdowns are making me feel funky, I feel like I should just use open models if others start playing at this game as well.
I was/am fine with Chinese companies offering cheaper subscriptions in China because ofc. But this is just weird, you could always charge me more let me try the damn thing, atleast it's own open router it seems like.
Initial impressions, interesting, it generated a pretty ok result for my prompt need to run a larger task but don't want to burn money so will wait for someone to burn their subscription and try running some opus 5.5 comparisons.
Would be great if Meta drops the weights, just to throw off Ant and OAI a bit
I don't think they'd have much to gain. People who use open-weights models will just use whatever open-weights model is the best. They'd lose a bunch of money, with little conversion to customers.
They've been saying they'll rise muse 1.2 for months now. Don't hold your breath
In the context of Meta trying to get non-technical people to embrace Muse agents to handle personal affairs, their internal metric showing a ~3% violation rate for mishandling personal information seems unacceptably high.
There seem to be no downsides to mishandling PI so I don't think it's a big concern to Meta
Reminded me of when George Hotz said only Zuck has “the button”. [0]
[0] https://youtu.be/-qRc7qou6XY?is=hWofxO0G6X3JjST5
spinning up a GPU and as many cores as possible?
mousewheel repurposed?
background follow your mouse?
neutral tones?
well then YOU are viewing a frontier model announcement!
While some are afraid that one rouge AI system may take over the world (or cause doom is some fashion), I think one of the best safety mechanisms to prevent this is having many competing and adversarial labs that are extremely close to each other. So it's good to see that Meta is not showing any intention of leaving the frontier.
It's interesting that in its marketing Meta compares Wonder to the closed frontier models but not the open (Chinese) models. By its own assessment, Meta could easily show how far ahead its own model is against the rest of the pack of open models but they chose not to. Why?
> Meta could easily show how far ahead its own model is against the rest of the pack of open models but they chose not to. Why?
Wonder is not an open model and there's no claims that it will be in their announcement.
Despite the claims that they would release weights for Muse Spark, they haven't done that yet either.
Does it weight tokens on their ability to elicit negative emotions in users in order to drive engagement?
Not that interesting unless it's made open weights. Frontier lasts weeks, inference is a commodity.
The only time I'll ever touch a Meta LLM is if it's powering the dystopian doom bot that is plunging its murder spike through the base of my skull.
Poetic
They added fancy annimations and effects to this specific webpage, which are unnecessary and distracting and lead to bad reading experience. I already suffer astigmatism and floaters. I hate it when I move my cursor and these gray blocks and characters flash and flicker.
I found it interesting in the example given of a user looking at house listings and the agent contacting the letting agent on it's own on behalf of the user.
Like the good ol' days of viruses and anti-viruses. I wonder if there's two sides that will play out; many agents across many users doing things at scale for their users, thereby creating an admin burden for someone else. However because the task is known there's a framework for how to solve the task, might Meta then provide an agent on the other side for the business owner to use.
Looks good, and at least it is available unlike Gemini 4 Argon
So, Wonder (and not Wonder Full) will "will soon power experiences in" their assistant app. But no rollout details .
Then, Full is meant for API usage and "Muse Code" only.
It's getting confusing what having "Muse" means.
The Mishandling Personal Information benchmark is non-sensical: that's the very definition of giving your personal data to Meta
Acquihires starting to finally pay off it seems. Bring on the competition
Have I been living under a rock, or this feels out of nowhere from Meta?
Seems they’ve just messed around with cheap and open weight models for years.
Then Muse appears, followed by full suite of frontier models…
they started their massive ai capex over a year ago with aquihiring scale ai, so I think it makes sense that their investment is paying off now. in total they've spent in the high 200 billion on ai, though I think most is likely on datacenters.
muse ai was the biggest consumer launch since chatgpt. That left them ( and lets them sit on top of ) tonnes of valuable exchanegs with people trying to get things done. That must be an amazing corpus for training.
Unlike everyone else, which at least has to pretend not to train on user data, they gave out their model for basically free if you agreed to let them train on it. Opencode shows nearly 40 trillion tokens going through muse spark 1.3 contributor this week alone. That almost definitely helped close the gap a lot.
oh boy, you can imagine the shit eating grin on the face of whoever came up with the name "wonder full".
RL team hill-climbing this to 2nd AA place and this fast is crazy
How does it work on day-to-day workflows? does it translates well?
I had to come back to this comment to make sure I wasn't dreaming... I remember seeing it at 2nd place this morning, but they have moved it to 5th, behind Gemini 4 Argon. Maybe it takes a while for the dust to settle around these benchmarks?
its three big companies at the frontier now? openai, anthropic and now meta.
why is google still not frontier?
They're not until they are. No doubt they are working on something and have better resourcing than the AI labs
Gemini 4 Argon release a few days ago was frontier a few days ago. Things move fast.
I think some of the benchmarks that Meta didn't pick show Gemini 4 Argon ahead but really the benchmarks are not that great any more and getting increasingly disconnected from reality as people search for "headroom" in the evals to avoid saturation I think. Opus 5.5 is "top" for example but I really dislike the way it talks to me (it feels quite condescending, holier-than-thou, uptight etc) and prefer Gemini 3.8 flash as a daily driver just because it's so much more pleasant (and faster!) to work with and it's code is good-enough for my needs, despite being weaker in the benchmarks.
Grok has been at the frontier since 4.6
I noticed Muse Wonder model on OpenRouter is age gated to 18+
If I'm reading this right, then at the moment, for certain tasks, this is the best model in the world.
A new company on the frontier. Not close, but actually on it. Wow this is big news.
This is amazing. I can’t wait benchmark to come out $ per performance
Apparently the benchmark-winning (behind Opus 5.5) Muse Wonder Full is only currently available to their subscribers.
Only Muse Wonder is available in the API.
Muse Glimmer, Muse Spark, Muse Wonder?
What's next, Sun Shine?
Shine, Shimmer, Glow, and ultimately, Cherenkov Radiation.
Meta's RL scaling seems to be in full force
Super interesting that they're not even making Wonder Full available outside of their own coding harness, and thus not providing an API price to use in comparisons.
impressive benchmark numbers… now how to actually use this without Zucking up my machine is a consideration.
The usage limits on their basic $5 code plan to try things out are ludicrously low?
I blew through 14% weekly usage and I'll I asked:
Beware...Man, I really hate meta stuff. Not only do they make it impossible to pay for running ads, but everytime I try and pay to try out Muse code I get an error message. I swear they would make 10x more money if they didn't make it impossible to pay them
The benchmarks are great for sure.
Now if only they could fix their coding harness so that it's usable for day-to-day use. I paid them for a month to try out Muse Code, but their sandbox never worked across Mac or Linux (both very stock installs, nothing special).
Muse Spark 1.3 is available in Cursor! You could try it there.
(Disclaimer: I work at Cursor/SpaceX)
I haven't tried it but there is the official `pi-muse-bridge` that can use your muse code subscription in pi.
Is this datacenter-only, or will they be open-sourcing?
The Kitchen game is a ripoff of Overcooked in pretty much everything lol
[dead]
This really does look compelling, but it would be very nice if they wouldn't test all of the stuff on Max!
I'm only interested which of the big techs is most generous when it comes to subsidizing consumer plans.
any idea if they will release the weights?
Low chance
Facebook successfully becoming a frontier lab was not on my bingo card for 2026, but the more competition the merrier I guess. We will have to see just how benchmaxxed this is but beating Astra and Fable is not easy.
What openai/anthropics answer to Muse app? Because I find it really compelling and with this new incredibly beefy model it will be even better.
OpenAI has Dots. I personally prefer the business model of subscription payments funding my AI assistant, instead of skimming a fee off the top whenever my AI assistant completes a purchase for me. The incentives are much more aligned with the user when the user is the customer. That said, based on the history of the web, Meta's approach will likely win out.
OpenAI has dots. Same concept. It’s less capable, but since I’m very OpenAI integrated I just had my dot install muse and use it
Claude is waiting for this to drop fable 5.5 lol.
Quite frankly, Meta's models have been so impressive (and cheap) that I haven't used an OpenAI or Claude model except for one-off tasks as of late. Interested to see where they go from here.
But when is another open weights release coming out...
Superin… you’ve got to be kidding me… slurp slurp little Zuck.
Beyond ridiculous
Just dont' use the word Super Intelligence
To be clear, Meta here is not using "Super Intelligence" the bullshit term Trump wants companies to adopt as a show of submission. They use "superintelligence" the bullshit term that Facebook made up for marketing reasons a few years ago. Both of these are of course entirely unrelated to the long-established meaning of the word (e.g. how Bostrom defines it).
Zuck has been using the term for a while in the context of Muse, if anything Trump took it from him
I know, it's... well, it's irritating to see Trump use it. But I could be wrong, I don't know the angle he's going for
I understand the general thing of it. SI might mean every person gets their personalized agent but you could already do this, but maybe the Chinese model works here, where they push the costs down
It's still stupid to start calling it all SI
The ML community has been using the phrase "Super Intelligence" long before Trump took office. https://en.wikipedia.org/wiki/Superintelligence:_Paths,_Dang... was published in 2014.
[dead]
how much benchmark-maxxing has been done here? taking bets on poly and kalshi...
I think it's safe to assume every single model is benchmaxxed. I mean why wouldn't you include training data from common benchmarks? Not only is a higher score good marketing, but these benchmarks are supposed to represent common tasks so training for those should in principle improve performance across a wider range of tasks.
There's an entire section in the linked post about measuring benchmaxxing, so I guess that probably says something.
All of them are doing all the benchmaxing they can most likely. Dumb not to when so many people care about those numbers and don’t have the opportunity to try every model out personally.
Muse Spark was generally good. I'd be surprised if this model felt benchmaxxed.