Desert Ant Labs: local, fast models that run on device(desertant.com) |
Desert Ant Labs: local, fast models that run on device(desertant.com) |
> Every model is free up to 100k monthly active devices. No tokens, no logins.
I dunno about the business model though. Cloud LLM billing makes sense: you're getting another computer to do work with each request, and using their compute via their gateway that they bill you.
These local models are like old school software. They're producing the weights, and then giving them to people. If I'm happy with the weights you've given me, and I'm not using your compute for inference and not wanting or needing any updates off you, why should you continue getting money off me and my customers?
The whole "but we need to keep it updated for your security" doesn't really work as well for software designed to run fully offline like these local models are.
I'm not saying they shouldn't get paid, but I guess I feel a personal sadness that it's less obvious how to successfully monetise such a sincerely useful and beneficial approach towards AI models.
I definitely understand the appeal of the desire to “buy it once”, but I think there are a few issues:
- almost no software is static. Look at a package like python Requests and even though it does the simplest thing and has barely changed from a user perspective, it gets updated all the time. This is true for most software. This is doubly true for something like local AI models where both the software and the hardware are changing constantly. Subscriptions motivate sellers to keep their software up to date.
- If I’m an app developer, the idea that I can try something out for $X/month is very appealing versus making an upfront investment of (let’s say) $X*20. This is doubly true for something like local models where I will almost certainly want something new when the models improve.
- To add to the first point, I work at a startup. No one asks questions when I want $20/month licenses. But let’s say I want something that’s gonna be in the 5 figure range annually. If I go to my CFO and ask for $50k upfront and then we implement something and the project fails, I look like an idiot. If I ask for $2000/mo budget for something and then we try it for two months and it fails, no one cares. Subscriptions are just safer in this sense.
- All compiled software is static. New versions are released. I'm happy to pay for new versions/features if I want/need them. But if the version I have is working offline, it doesn't need security updates, and if it does what I need, it doesn't need new features. Same applies if I'm building software for others, if the library does what I need for my software for my customers and I need nothing more off the dev, why should I keep paying the dev?
- I'm struggling not be sarcastic here, sorry, but this might shock you to hear: most humans find the idea of being able to try whatever they want and get paid for it appealing. That's not really how money or life works though.
- I don't know how to respond to this point, it's a bit of a strawman. Yes, if the only other option is to pay $50k up front for a local model, then your point is valid.
As a kid, I spent literally hundreds of £ on the first Sims, and Sims expansion packs at £20 /pop. EA got loads of money off me, I got loads of new features I wanted. I moaned at the time, but on reflection, everyone was a winner. Sims is still going now (possibly as a subscription!)
Again, apologies for a bit of sass here (that's a me problem), but the number of people who seem to sincerely believe the points you've put forward is why I'm personally sad at the state of modern software economics.
Everyone wants to be paid forever for something they produced once is some kind of a mind virus. Make me a better chair and maybe I will buy it, but don't expect to become a trillion dollar company. It's deeply unfair to everyone who wants to be a trillion dollar company of course.
What is really funny to me -- the ones that do make it to collect the rent indefinitely also decrease their own taxes paid to the government AND also decrease the amount of contribution to society by making less and shittier stuff.
You are, presumably, making copies of their software for your customers. So I guess the real question here is, should copyright exist? For software?
This is actually a classic enterprise software approach - as there are no fixed per unit production costs to cover, the pricing is very elastic.
ie each license is probably a negotiation - if you see your app for a one off fee then they want a one-off cut - if you sell a subscription they may want a cut of that instead.
Obviously the risk is if you bake it in and your app becomes successful they may want a larger cut than you can afford, but if they don't negotiate in good faith they won't last very long.
Feels like it'd be more aligned with how the business and users would produce and use it, it'd be something like: "We've done a bunch of research and output, here is the thing, you can buy it, and we'll work on the next, then you can pay again for the next if you'd like".
If neither of those are true, a subscription or recurring license or similar regular fee just feels like greed/robbery.
Because they own the IP and they get to decide the terms of how it’s licensed.
This is like a company taking open source software, saying they like the code the community has given them, and asking why they should continue having to respect the terms of the license after downloading the code. The availability of the software (or models) does not equal a free license to use as you please.
A license allowing 100K devices for free is very generous. The businesses selling more than 100K units of anything will be significant operations. It’s fair that they’re asked to contribute financially.
how'd that work out for the rest of IP owners that had their stuff used as training data for use in model creation?
was the material they trained on produced in-house?
Well, the laws that society decide what is legal in terms of buying and selling.
People, somewhat unfortunately, are starting to act as if "licenses" are some kind of natural, pre-existing part of the world.
The fact that they exist at all is because we invented them, and we can easily uninvent them.
You made up a position to argue against.
What do you mean by "continue" if you haven't paid anything yet at that point?
Welcome to the concept of fair market value. Less snarkily you're conflating the concepts of price and cost; theyre not the same thing and theyre not the same for you or the seller.
Because that's the contract you agreed to?
If you don't like the contract they offer, don't take it. You can choose something else. It's really as simple as that.
This. I run small models (>50MB) for bio-imaging/biotech applications, it feels like every README implies that you need a discrete GPU to get started. While some do, many, especially the most useful ones, do not. Sure it matters if you're also going to do fine-tuning, but I believe your typical user just wants to detect some nuclei and get some cell-body ratios.
The laptop on your desk won't be running Meta's SAM, but it has more than enough compute to crunch 100's of your H&E slides overnight.
[0] https://github.com/mouseland/cellpose
>accessible via one SDK for Swift, Kotlin, and JavaScript
Lol well let me know when there's a Python SDK and I'll give it a try then. Obviously this isn't a deal breaker if you have a real case, but as someone who is willing to spin something up and try it out if there's a quick "pip install" command, this is getting put back on the shelf for now.
Look forward to the next model
> How much does Voz cost?
> Every model is free up to 100k monthly active devices per SDK. Unlimited inference per user. Contact us for custom licenses.
I tried out their demo for Clear, the audio quality improvement model.[0] I'm not sure if it's just I don't have refined enough an ear or their demo is broken, but the "raw" and "enhanced" versions sounded exactly the same to me.
Nvidia's RE-USE model can do what MossFormer2 does _and_ can remove reverb, but it is non-commercial licensed.
If any of you are interested there are apps like https://github.com/altic-dev/FluidVoice (this one's a great app) and this https://sam-pop.github.io/WhisperDictation. The latter, even though it has just 7 stars right now, seems to be more "intuitive". I just hope they expose a way to "connect" to available models on the machine or remotely)
Voz is Parakeet 0.6B v3
Clear is DeepFilterNet 3
Ear is the language predictor from whisper-tiny
...
Most organisations give their users terrible hardware, so anything which requires 32GB of RAM or a MacBook Pro won’t work if it’s a government or large organisation.
This is very impressive for a 248MB model. I wonder how good the results are, as an LLM 10x the size is still quite bad at that.
> Hate speech triage. On-device moderation that flags hateful, abusive and threatening text
What could go wrong here?
This also could be ideal if you want your child to play online to have auto-censorship
2. Having human in the loop to review it and subject said human to the worst other humans produce.
Once image support drops this could be a cool addition to https://struktur.sh. Will it be on OpenRouter too or will all inference have to be self-managed? I’m thinking about server-side use cases, where budgets are low, so small models shine.
edit: ah I saw mainly iOS for now. But an integration would be possible on macOS then, right?
So I think it's some buggy code in their UI that even prevents the model from running.
That's a problem with the whole AI space - it's eating itself at a incredibly rate.
Ultimately if they add value and behave well they might carve out a market - but anything in the software space right now is a gamble.
A 100K free device license is a generous gift to small companies. Past 100K, it’s more than fair to ask that companies contribute. That’s a significant operation at that point and the model has obviously provided value.
If the model doesn’t provide any value to a >100K device company then they should train their own or not use it.
I could see how people unfamiliar with per-device or per-unit licensing fees would be confused, but this is a common business model. Scaling payments with the customer’s business success is one of the more fair ways to align vendor and customer while also providing a nice way to give freely to smaller businesses like this.
It’s a good thing. It’s silly that a company giving away a license for 100K devices, which covers small and medium hardware operations, is being criticized for this of all things.
We aim to make all our models available cross-platform. There's a subset currently only iOS/macOS, because they take some more effort to port over to Android and web.
Check out the CLI for easy experimentation: https://github.com/Desert-Ant-Labs/desert-ant-cli#desert-ant...
I know some people want it to be a legal violation, but it’s not.
Also these are audio models, if you hadn’t noticed. The training sets for this type of work is very different than the corpus of scraped GitHub repos.
It will take years for the hundreds of jurisdictions in the world to write new laws and (in common law countries) establish new precedent.
Philosophical questions don't make sense if you use them to counter argue about something which is a legal issue.
Not only did you pull a whataboutism but the whataboutism isn't even relevant to the topic of conversation lol.
Here is to more beauty in the world - both physical & digital :)
You’re talking about times before 2010 and an EA game, but that’s not the right comparison. These Desert Ant AI models are an SDK, a professional middleware library. Licensing that kind of software in the pre-2010 days used to cost tens of thousands of dollars.
In other words, the proper comparison here is the various libraries that EA licensed to make the game, not the game itself.
> I don't know how to respond to this point, it's a bit of a strawman.
Why is it a strawman? I can sell you software for $100 once or I can sell it to you for $10/month. I don't think anyone is suggesting that I'm gonna sell it for $10 once though.
In particular, Desert Ant is saying their software is free for 100k active devices per SDK. That's likely a lot more users than most peoples' side projects. Given that, I'm guessing that $50k (even annually) is not a crazy number (yes, obviously $50k is not what a consumer would ever pay).
As in - I make software for a living and they can fuck off. Produce the item. Sell the item at a cost you find reasonable. Convince me to buy it again with an upgrade that's worth buying.
Period.
There's no "service" happening here. If anything, the proposed cost model just makes me incredibly unlikely to engage at all, because it means they're going to insist on trying to track my users (a thing I don't like doing...) with whatever bloatware they shove into their "sdk".
They're selling it to your employer. You're literally not even the customer for this product. There's no need to get angry that they're charging enterprise/vendor rates for a product.
And it's free for up to 100k devices. If you are running anything on 100k devices you should be able to afford $0.50/device and. If you aren't then why do you care? It's still a very customer-permissive and flexible business model so I cannot understand why this is upsetting or offensive
An alternative would be instead of 1 of charging over 100k users $50k, charge every user a one off $20.
I guess it's hard to discuss alternative billing methods without knowing what need to be recuperated.
Thanks for the response, appreciated. Again, sorry for the sassyness.
Who is allowed to make copies of the library?
I deploy browser extensions to millions of devices, and small local models that I can run in that space are incredibly appealing.
But the "per user" model is just especially difficult to justify here, and comes with a TON of fucking downsides like usage tracking, feature availability, and product structure (ex - is a trial user a user? how long? etc...)
---
And it's not even like they're justifying it with model download support or caching... they shove it entirely onto hugging face, or have you self serve.
As their literal target audience - I'd rather buy the weights for a fixed price (even a high one), and not let them influence my product design decisions with saas style pricing.
Then make a better chair yourself. Also, I don't see how chair2 needs r&d and incurs hardware costs.
That is very reasonable.
Seriously, look around you, pick a thing, imagine a world where the price you paid for that thing hinged on your personal individual success with that thing.
edit: to clarify, my point is why can't we pay for the model weights up front for offline models? Why would we continue paying for things we don't want/need anymore, like updates for offline software that does it's job fine?
Unless you could lease a chair and limit your cost to some finite number (which is higher then upfront cost), because chairs are easy to get by. You could also buy very cheap chairs from China for example and rent them by the hour (or that's what I heard).
The actual reason to why is -- because they can make you do it.
>What other things work like that outside of the modern tech world?
taxes
YCombinator. Unreal Engine. CryEngine. FMOD. Steam, App stores in general. Qt.
+ 90% of enterprise software that charges per seat / user.
You could argue that most of these provide ongoing services, but that will also be the case if you license one of their models commercially, contracts come with strings attached both ways.
I don't think many chair-users would opt for it, but I see no reason to ban it. Some services already work this way, like accounting firms that specialise in doing self-employed people's taxes.