formulas not rendered
In the iOS version of the Locally app, mathematical formulas are not rendered correctly.
Instead of displaying formatted equations, the app shows the raw LaTeX source code. When I copy this LaTeX code and paste it into any external editor that supports LaTeX/MathML or Markdown math, the formulas render correctly, which suggests the issue is in the client-side rendering, not in the generated content.
Instead of displaying formatted equations, the app shows the raw LaTeX source code. When I copy this LaTeX code and paste it into any external editor that supports LaTeX/MathML or Markdown math, the formulas render correctly, which suggests the issue is in the client-side rendering, not in the generated content.
Show more
Good start!
It’s a nice start for dead-easy to use App Store version of LM Studio including some new LLMs. However, it doesn’t include newest versions, and it seems clear that the ones selected are made to work on iPad as the ceiling. This is a Mac app - let us kit out with LLMs that use our Pro silicon!
Very good! Missing some features
The app is great and does what it promises. However, there are some “quality of life” things I’ve come to take for granted from LM Studio that aren’t currently supported here, such as editing prompts, responses, or thoughts. Likewise, you can only have one personalization prompt. All the same, it’s cool that everything is so mindful of privacy.
Show more
Nice, but clunky
It technically does work and runs LLMs on your iPhone. A major gripe I have with it though is that LaTeX expressions are FREQUENTLY displayed incorrectly. That is, the model is writing valid LaTeX but Locally’s renderer doesn’t display it properly. Before the latest update it was shaky—long expressions couldn’t be scrolled at all, and some shorter ones had weird layout issues and looked like they were off-place a couple dozen pixels—but after the latest update, all LaTeX fields just appear blank entirely. I would’ve given this app 5 stars if it weren’t for LaTeX issues.
Also, I highly recommend adding some kind of web search option like Ollama Web Search. That’d be mighty useful for letting the model know about current events and info it may not be able to come up with on its own.
Also, I highly recommend adding some kind of web search option like Ollama Web Search. That’d be mighty useful for letting the model know about current events and info it may not be able to come up with on its own.
Show more
New good models don't support
IPhone 17 pro max
Bonsai 27B - doesn't work after 2nd message.
Gemma 4 E2B - doesn't support at all.
Bonsai 27B - doesn't work after 2nd message.
Gemma 4 E2B - doesn't support at all.
Best free app
Love regular updates
Garbage ai models
Argued with an 8B model about how old i was. Can’t do elementary level math since it didn’t account for the months and days, when i corrected it, it showed me the math and still provided the wrong answer 💀
Pretty good
App generates fast which is a big plus. I wish it would let you import models from huggingface. The choice the app gives you is rather small and highly censored.
LM Link is clutch
Being able to chat with a model running on my desktop is the best feature. Unfortunately, it doesn't seem to load the same model config as the desktop setup. Really need to include the setting for continuing the session beyond the context window, instead of just throwing an error. Also, it'd be great to share/sync conversations between devices.
Show more
doesn’t work
doesn’t work and crashes
Crashes with Bonsai 27b
Numerous crashes with new Bonsai 27b release: 1. Opening app without chatting. 2. Basic prompt and crashes during response. 3. Coding prompt and fails to finish response, then crashes when viewing the partially provided code response. Unable to prompt in that same chat.
Device: iPhone 17 Pro Max
iOS: 26.4.1
Device: iPhone 17 Pro Max
iOS: 26.4.1
Show more
Great solution for Apple Shortcuts
Best option I’ve found for automating LLM interactions through Apple Shortcuts; the Locally AI app provides a nice GUI for managing local processing models, and the Shortcuts action gives you text and file attachment options.
There’s no need to keep the app running, Shortcuts appears to call it invisibly in the background, and you can even set up model selection as part of the shortcut setup procedure. Fantastic!
There are a few frustrations/bugs that I hope get some attention soon:
1. It’s only possible to download one model at a time, and you cannot queue downloads (so when you want to add 4 models, you have to wait for each one to finish before starting the next)
2. Gemma 4 E4B and E2B models are currently broken and fail to load (I’ve sent a bug report via email)
But overall, this is the best I’ve found so far, and I offer my thanks to the LMStudio team for making this available, for free no less!
There’s no need to keep the app running, Shortcuts appears to call it invisibly in the background, and you can even set up model selection as part of the shortcut setup procedure. Fantastic!
There are a few frustrations/bugs that I hope get some attention soon:
1. It’s only possible to download one model at a time, and you cannot queue downloads (so when you want to add 4 models, you have to wait for each one to finish before starting the next)
2. Gemma 4 E4B and E2B models are currently broken and fail to load (I’ve sent a bug report via email)
But overall, this is the best I’ve found so far, and I offer my thanks to the LMStudio team for making this available, for free no less!
Show more
Very good app with clean UI
I really enjoy using this app. It has a very clean and easy understandable UI. It offers a rather limited set of functionality, but that is actually a positive. Instead of overloading the UI with bloated half-baked functionality this one is clean and straight forward and what is there actually works. You can start testing local AI models in minutes without much effort. Well done!
Gave it four stars and encourage you to improve it even more. Like give us control over the context length to be used. It seems limiting at the moment compared to Ollama with the same model. Could also show some more info or statistics (tokens/sec, context, memory etc.) while it is running.
Looking forward to the next update. Keep up the good work!
Gave it four stars and encourage you to improve it even more. Like give us control over the context length to be used. It seems limiting at the moment compared to Ollama with the same model. Could also show some more info or statistics (tokens/sec, context, memory etc.) while it is running.
Looking forward to the next update. Keep up the good work!
Show more
Need support for MCP
There are mcp working with LM studio but can’t let it use with this app such as search_web. Make it available for the app as well. Thanks.
No recommended
Garbage, I downloaded gemma4 model, tried to send a prompt it wouldn’t let me, the send button was greyed out, I could only use apple foundation model
Actually works
Many local LLM apps exist but don’t work. This does. Also there is a better model selection built into this app than I have found on alternatives; and even a way to download from your own source.
Yes thank y’all so much
The common man needs local AI governance. YES!!!!!
LM link is fantastic
I’ve enjoyed using Locally on my iPad for a while now. The recent addition of LM Link is amazing. The ability to serve much larger models with a Mac is exactly what is needed. Maintaining privacy while being able to use bigger models is perfect. So far this feature has been working very smoothly with my Mac Studio Max M4 and my iPad and iPhone. The Mac running LM Studio and the iPad running Locally AI.
A big thanks to the developers.
A big thanks to the developers.
Show more
can not download model
In China, I tried in normal internet and VPN. Both can’t work.
It’s okay
I can’t choose model parameters, download new models to my machine, or use MCP servers
