Just ordered a $12k mac studio w/ 512GB of integrated RAM.
Can't wait for it to arrive and crank up LM Studio. It's literally the first install. I'm going to download it with safari.
LM Studio is newish, and it's not a perfect interface yet, but it's fantastic at what it does which is bring local LLMs to the masses w/o them having to know much.
Exo is this radically cool tool that automatically clusters all hosts on your network running Exo and uses their combined GPUs for increased throughput.
Like HPC environments, you are going to need ultra fast interconnects, but it's just IP based.
If the primary use case is input heavy, which is true of agentic tools, there’s a world where partial GPU offload with many channels of DDR5 system RAM leads to an overall better experience. A good GPU will process input many times faster, and with good RAM you might end up with decent output speed still. Seems like that would come in close to $12k?
And there would be no competition for models that do fit entirely inside that VRAM, for example Qwen3 32B.
I'm using it on MacBook Air M1 / 8 GB RAM with Qwen3-4B to generate summaries and tags for my vibe-coded Bloomberg Terminal-style RSS reader :-) It works fine (the laptop gets hot and slow, but fine).
Probably should just use llama.cpp server/ollama and not waste a gig of memory on Electron, but I like GUIs.
8 GB of RAM with local LLMs in general is iffy: a 8-bit quantized Qwen3-4B is 4.2GB on disk and likely more in memory. 16 GB is usually the minimum to be able to run decent models without compromising on heavy quantization.
I'd love to host my own LLMs but I keep getting held back from the quality and affordability of Cloud LLMs. Why go local unless there's private data involved?
Not OP, but with LM Studio I get a chat interface out-of-the-box for local models, while with openwebui I'd need to configure it to point to an OpenAI API-compatible server (like LM Studio). It can also help determine which models will work well with your hardware.
LM Studio isn't FOSS though.
I did enjoy hooking up OpenWebUI to Firefox's experimental AI Chatbot. (browser.ml.chat.hideLocalhost to false, browser.ml.chat.provider to localhost:${openwebui-port})
i recently tried openwebui but it was so painful to get it to run with local model.
that "first run experience" of lm studio is pretty fire in comparison. can't really talk about actually working with it though, still waiting for the 8GB download
Yup, I'm spoiled by Claude 3.7 Sonnet right now. I had to stop using opus for plan mode in my Agent because it is just so expensive. I'm using Gemini 2.5 pro for that now.
I was using Claude 3.7 exclusively for coding, but it sure seems like it got worse suddenly about 2–3 weeks back. It went from writing pretty solid code I had to make only minor changes to, to being completely off its rails, altering files unrelated to my prompt, undoing fixes from the same conversation, reinventing db access and ignoring existing coding 'standards' established in the existing codebase. Became so untrustworthy I finally gave OpenAi O3 a try and honestly, I was pretty surprised how solid it has been. I've been using o3 since, and I find it generally does exactly what I ask, esp if you have a well established project with plenty of code for it to reference.
Just wondering if Claude 3.7 has seemed differently lately for anyone else? Was my go to for several months, and I'm no fan of OpenAI, but o3 has been rock solid.
The initial experience with LMStudio and MCP doesn't seem to be great, I think their docs could do with a happy path demo for newcomers.
Upon installing the first model offered is google/gemma-3-12b - which in fairness is pretty decent compared to others.
It's not obvious how to show the right sidebar they're talking about, it's the flask icon which turns into a collapse icon when you click it.
I set the MCP up with playwright, asked it to read the top headline from HN and it got stuck on an infinite loop of navigating to Hacker News, but doing nothing with the output.
I wanted to try it out with a few other models, but figuring out how to download new models isn't obvious either, it turned out to be the search icon. Anyway other models didn't fare much better either, some outright ignored the tools despite having the capacity for 'tool use'.
Gemma3 models can follow instructions but were not trained to call tools, which is the backbone of MCP support. You would likely have a better experience with models from the Qwen3 family.
Great to see more local AI tools supporting MCP! Recently I've also added MCP support to recurse.chat. When running locally (LLaMA.cpp and Ollama) it still needs to catch up in terms of tool calling capabilities (for example tool call accuracy / parallel tool calls) compared to the well known providers but it's starting to get pretty usable.
claude going mcp over remote kinda normalised the protocol for inference routing. now with lmstudio running as local mcp host, you can just tunnel it (cloudflared/ngrok), drop a tiny gateway script and boom your laptop basically acts like a mcp node in hybrid mesh. short prompts hit qwen local, heavier ones go claude. with same payload and interface we can actually get multihost local inference clusters wired together by mcp
What models are you using on LM Studio for what task and with how much memory?
I have a 48GB macbook pro and Gemma3 (one of the abliterated ones) fits my non-code use case perfectly (generating crime stories which the reader tries to guess the killer).
LM Studio has quickly become the best way to run local LLMs on an Apple Silicon Mac: no offense to vllm/ollama and other terminal-based approaches, but LLMs have many levers for tweaking output and sometimes you need a UI to manage it. Now that LM Studio supports MLX models, it's one of the most efficient too.
I'm not bullish on MCP, but at the least this approach gives a good way to experiment with it for free.
This isn't true. You can `ollama run {model}`, `/set parameter num_ctx {ctx}` and then `/save`. Recommended to `/save {model}:{ctx}` to persist on model update
I just wish they did some facelifting of UI. Right now is too colorfull for me and many different shades of similar colors. I wish they copy some color pallet from google ai studio or from trae or pycharm.
Near as I can tell it's supposed to make calling other people's tools easier. But I don't want to spin up an entire server to invoke a calculator. So far it seems to make building my own local tools harder, unless there's some guidebook I'm missing.
It's a protocol that doesn't dictate how you are calling the tool. You can use in-memory transport without needing to spin up a server. Your tool can just be a function, but with the flexibility of serving to other clients.
I wish LM Studio had a pure daemon mode. It's better than ollama in a lot of ways but I'd rather be able to use BoltAI as the UI, as well as use it from Zed and VSCode and aider.
What I like about ollama is that it provides a self-hosted AI provider that can be used by a variety of things. LM Studio has that too, but you have to have the whole big chonky Electron UI running. Its UI is powerful but a lot less nice than e.g. BoltAI for casual use.
Oh, that horrible Electron UI. Under Windows it pegs a core on my CPU at all times!
If you're just working as a single user via the OpenAI protocol, you might want to consider koboldcpp. It bundles a GUI launcher, then starts in text-only mode. You can also tell it to just run a saved configuration, bypassing the GUI; I've successfully run it as a system service on Windows using nssm.
Though there are a lot of roleplay-centric gimmicks in its feature set, its context-shifting feature is singular. It caches the intermediate state used by your last query, extending it to build the next one. As a result you save on generation time with large contexts, and also any conversation that has been pushed out of the context window still indirectly influences the current exchange.
Still, you need to install and run the AppImage at least once to enable the "lms" cli which can later be used. Would be nice with a completely GUI-less installation/use method too.
Can't wait for it to arrive and crank up LM Studio. It's literally the first install. I'm going to download it with safari.
LM Studio is newish, and it's not a perfect interface yet, but it's fantastic at what it does which is bring local LLMs to the masses w/o them having to know much.
There is another project that people should be aware of: https://github.com/exo-explore/exo
Exo is this radically cool tool that automatically clusters all hosts on your network running Exo and uses their combined GPUs for increased throughput.
Like HPC environments, you are going to need ultra fast interconnects, but it's just IP based.
Get the RTX Pro 6000 for 8.5k with double the bandwidth. It will be way better
If the primary use case is input heavy, which is true of agentic tools, there’s a world where partial GPU offload with many channels of DDR5 system RAM leads to an overall better experience. A good GPU will process input many times faster, and with good RAM you might end up with decent output speed still. Seems like that would come in close to $12k?
And there would be no competition for models that do fit entirely inside that VRAM, for example Qwen3 32B.
The whole point of spending that much money for them is to run massive models, like the full R1, which the Pro 6000 cant
Probably should just use llama.cpp server/ollama and not waste a gig of memory on Electron, but I like GUIs.
https://www.pcgamer.com/apple-vp-says-8gb-ram-on-a-macbook-p...
Oof you were NOT joking
LM Studio isn't FOSS though.
I did enjoy hooking up OpenWebUI to Firefox's experimental AI Chatbot. (browser.ml.chat.hideLocalhost to false, browser.ml.chat.provider to localhost:${openwebui-port})
I'm interested in using models for code generation, but I'm not expecting much in that regard.
I'm planning to attempt fine tuning open source models on certain tool sets, especially MCP tools.
I haven’t been using it much. All it has on it is LM Studio, Ollama, and Stats.app.
> Can't wait for it to arrive and crank up LM Studio. It's literally the first install. I'm going to download it with safari.
lol, yup. same.
I'm considering ordering one of these today: https://www.newegg.com/p/N82E16816139451?Item=N82E1681613945...
It looks like it will hold 5 GPUs with a single slot open for infiniband
Then local models might be lower quality, but it won't be slow! :)
Just wondering if Claude 3.7 has seemed differently lately for anyone else? Was my go to for several months, and I'm no fan of OpenAI, but o3 has been rock solid.
Upon installing the first model offered is google/gemma-3-12b - which in fairness is pretty decent compared to others.
It's not obvious how to show the right sidebar they're talking about, it's the flask icon which turns into a collapse icon when you click it.
I set the MCP up with playwright, asked it to read the top headline from HN and it got stuck on an infinite loop of navigating to Hacker News, but doing nothing with the output.
I wanted to try it out with a few other models, but figuring out how to download new models isn't obvious either, it turned out to be the search icon. Anyway other models didn't fare much better either, some outright ignored the tools despite having the capacity for 'tool use'.
I'd love to learn more about your MCP implementation. Wanna chat?
I have a 48GB macbook pro and Gemma3 (one of the abliterated ones) fits my non-code use case perfectly (generating crime stories which the reader tries to guess the killer).
For code, I still call Google to use Gemini.
Nice to have a local option, especially for some prompts.
I'm not bullish on MCP, but at the least this approach gives a good way to experiment with it for free.
You gotta help me out. What do you see holding it back?
What I like about ollama is that it provides a self-hosted AI provider that can be used by a variety of things. LM Studio has that too, but you have to have the whole big chonky Electron UI running. Its UI is powerful but a lot less nice than e.g. BoltAI for casual use.
If you're just working as a single user via the OpenAI protocol, you might want to consider koboldcpp. It bundles a GUI launcher, then starts in text-only mode. You can also tell it to just run a saved configuration, bypassing the GUI; I've successfully run it as a system service on Windows using nssm.
https://github.com/LostRuins/koboldcpp/releases
Though there are a lot of roleplay-centric gimmicks in its feature set, its context-shifting feature is singular. It caches the intermediate state used by your last query, extending it to build the next one. As a result you save on generation time with large contexts, and also any conversation that has been pushed out of the context window still indirectly influences the current exchange.
Are you sharing any of your revenue from that $79 license fee with the https://ollama.com/ project that your app builds on top of?