Can I run AI locally?

(canirun.ai)

223 points | by ricardbejarano 5 hours ago ago

53 comments

  • meatmanek 18 minutes ago

    This seems to be estimating based on memory bandwidth / size of model, which is a really good estimate for dense models, but MoE models like GPT-OSS-20b don't involve the entire model for every token, so they can produce more tokens/second on the same hardware. GPT-OSS-20B has 3.6B active parameters, so it should perform similarly to a 3-4B dense model, while requiring enough VRAM to fit the whole 20B model.

    (In terms of intelligence, they tend to score similarly to a dense model that's as big as the geometric mean of the full model size and the active parameters, i.e. for GPT-OSS-20B, it's roughly as smart as a sqrt(20b*3.6b) ≈ 8.5b dense model, but produces tokens 2x faster.)

  • sdingi 2 minutes ago

    When running models on my phone - either through the web browser or via an app - is there any chance it uses the phone's NPU, or will these be GPU only?

    I don't really understand how the interface to the NPU chip looks from the perspective of a non-system caller, if it exists at all. This is a Samsung device but I am wondering about the general principle.

  • tcbrah 3 minutes ago

    tbh i stopped caring about "can i run X locally" a while ago. for anything where quality matters (scripting, code, complex reasoning) the local models are just not there yet compared to API. where local shines is specific narrow tasks - TTS, embeddings, whisper for STT, stuff like that. trying to run a 70b model at 3 tok/s on your gaming GPU when you could just hit an API for like $0.002/req feels like a weird flex IMO

  • twampss an hour ago

    Is this just llmfit but a web version of it?

    https://github.com/AlexsJones/llmfit

    • deanc an hour ago

      Yes. But llmfit is far more useful as it detects your system resources.

      • dgrin91 an hour ago

        Honestly I was surprised about this. It accurately got my GPU and specs without asking for any permissions. I didnt realize I was exposing this info.

        • dekhn an hour ago

          How could it not? That information is always available to userspace.

          • bityard 6 minutes ago

            "Available to userspace" is a much different thing than "available to every website that wants it, even in private mode".

            I too was a little surprised by this. My browser (Vivladi) makes a big deal about how privacy-conscious they are, but apparently browser fingerprinting is not on their radar.

            • swiftcoder a minute ago

              It's pretty hard to avoid GPU fingerprinting if you have webgl/webgpu enabled

            • dekhn a minute ago

              We switched to talking about llmfit in this subthread, it runs as native code.

        • rithdmc 33 minutes ago

          Do you mean the OPs website? Mine's way off.

          > Estimates based on browser APIs. Actual specs may vary

  • LeifCarrotson an hour ago

    This lacks a whole lot of mobile GPUs. It also does not understand that you can share CPU memory with the GPU, or perform various KV cache offloading strategies to work around memory limits.

    It says I have an Arc 750 with 2 GB of shared RAM, because that's the GPU that renders my browser...but I actually have an RTX1000 Ada with 6 GB of GDDR6. It's kind of like an RTX 4050 (not listed in the dropdowns) with lower thermal limits. I also have 64 GB of LPDDR5 main memory.

    It works - Qwen3 Coder Next, Devstral Small, Qwen3.5 4B, and others can run locally on my laptop in near real-time. They're not quite as good as the latest models, and I've tried some bigger ones (up to 24GB, it produces tokens about half as fast as I can type...which is disappointingly slow) that are slower but smarter.

    But I don't run out of tokens.

  • carra 35 minutes ago

    Having the rating of how well the model will run for you is cool. I miss to also have some rating of the model capabilities (even if this is tricky). There are way too many to choose. And just looking at the parameter number or the used memory is not always a good indication of actual performance.

  • sxates 2 hours ago

    Cool thing!

    A couple suggestions:

    1. I have an M3 Ultra with 256GB of memory, but the options list only goes up to 192GB. The M3 Ultra supports up to 512GB. 2. It'd be great if I could flip this around and choose a model, and then see the performance for all the different processors. Would help making buying decisions!

  • Felixbot an hour ago

    The RAM/VRAM cutoff matters more than the parameter count alone. A 13B model in Q4_K_M quantization fits in 8GB VRAM with reasonable throughput, but the same model in fp16 needs 26GB. Most calculators treat quantization as a footnote when it is actually the primary variable. The question is not "can I run 13B" but "what quantization level gives acceptable quality at my hardware ceiling".

  • freediddy 18 minutes ago

    i think the perplexity is more important than tokens per second. tokens per second is relatively useless in my opinion. there is nothing worse than getting bad results returned to you very quickly and confidently.

    ive been working with quite a few open weight models for the last year and especially for things like images, models from 6 months would return garbage data quickly, but these days qwen 3.5 is incredible, even the 9b model.

    • sroussey 14 minutes ago

      No, getting bad results slowly is much worse. Bad results quickly and you can make adjustments.

      But yes, if there is a choice I want quality over speed. At same quality, I definitely want speed.

  • phelm an hour ago

    This is awesome, it would be great to cross reference some intelligence benchmarks so that I can understand the trade off between RAM consumption, token rate and how good the model is

  • orthoxerox 24 minutes ago

    For some reason it doesn't react to changing the RAM amount in the combo box at the top. If I open this on my Ryzen AI Max 395+ with 32 GB of unified memory, it thinks nothing will fit because I've set it up to reserve 512MB of RAM for the GPU.

  • amelius 11 minutes ago

    What is this S/A/B/C/etc. ranking? Is anyone else using it?

    • vikramkr 10 minutes ago

      Just a tier list I think

    • relaxing 6 minutes ago

      Apparently S being a level above A comes from Japanese grading. I’ve been confused by that, too.

  • GrayShade 2 hours ago

    This feels a bit pessimistic. Qwen 3.5 35B-A3B runs at 38 t/s tg with llama.cpp (mmap enabled) on my Radeon 6800 XT.

    • Aurornis 30 minutes ago

      At what quantization and with what size context window?

  • amelius 12 minutes ago

    Why isn't there some kind of benchmark score in the list?

  • sshagent an hour ago

    I don't see my beloved 5060ti. looks great though

  • AstroBen 33 minutes ago

    This doesn't look accurate to me. I have an RX9070 and I've been messing around with Qwen 3.5 35B-A3B. According to this site I can't even run it, yet I'm getting 32tok/s ^.-

    • misnome 18 minutes ago

      It seems to be missing a whole load of the quantized Qwen models, Qwen3.5:122b works fine in the 96GB GH200 (a machine that is also missing here....)

  • John23832 2 hours ago

    RTX Pro 6000 is a glaring omission.

    • embedding-shape an hour ago

      Yeah, that's weird, seems it has later models, and earlier, but specifically not Pro 6000? Also, based on my experience, the given numbers seems to be at least one magnitude off, which seems like a lot, when I use the approx values for a Pro 6000 (96GB VRAM + 1792 GB/s)

    • schaefer 2 hours ago

      No Nvidia Spark workstation is another omission.

  • ge96 an hour ago

    Raspberry pi? Say 4B with 4GB of ram.

    I also want to run vision like Yocto and basic LLM with TTS/STT

    • boutell 19 minutes ago

      I've been trying to get speech to text to work with a reasonable vocabulary on pis for a while. It's tough. All the modern models just need more GPU than is available

      • ge96 14 minutes ago

        Whispr?

        For wakewords I have used pico rhino voice

        I want to use these I2S breakout mics

  • vova_hn2 an hour ago

    It says "RAM - unknown", but doesn't give me an option to specify how much RAM I have. Why?

  • mrdependable an hour ago

    This is great, I've been trying to figure this stuff out recently.

    One thing I do wonder is what sort of solutions there are for running your own model, but using it from a different machine. I don't necessarily want to run the model on the machine I'm also working from.

  • jrmg 34 minutes ago

    Is there a reliable guide somewhere to setting up local AI for coding (please don’t say ‘just Google it’ - that just results in a morass of AI slop/SEO pages with out of date, non-self-consistent, incorrect or impossible instructions).

    I’d like to be able to use a local model (which one?) to power Copilot in vscode, and run coding agent(s) (not general purpose OpenClaw-like agents) on my M2 MacBook. I know it’ll be slow.

    I suspect this is actually fairly easy to set up - if you know how.

    • AstroBen 24 minutes ago

      Ollama or LM Studio are very simple to setup.

      You're probably not going to get anything working well as an agent on an M2 MacBook, but smaller models do surprisingly well for focused autocomplete. Maybe the Qwen3.5 9B model would run decently on your system?

  • havaloc an hour ago

    Missing the A18 Neo! :)

  • debatem1 an hour ago

    For me the "can run" filter says "S/A/B" but lists S, A, B, and C and the "tight fit" filter says "C/D" but lists F.

    Just FYI.

  • adithyassekhar an hour ago

    This just reminded me of this https://www.systemrequirementslab.com/cyri.

    Not sure if it still works.

  • arjie an hour ago

    Cool website. The one that I'd really like to see there is the RTX 6000 Pro Blackwell 96 GB, though.

  • nilslindemann 14 minutes ago

    1. More title attributes please ("S 16 A 7 B 7 C 0 D 4 F 34", huh?)

    2. Add a 150% size bonus to your site.

    Otherwise, cool site, bookmarked.

  • kylehotchkiss 20 minutes ago

    My Mac mini rocks qwen2.5 14b at a lightning fast 11/tokens a second. Which is actually good enough for the long term data processing I make it spend all day doing. It doesn’t lock up the machine or prevent its primary purpose as webserver from being fulfilled.

  • varispeed 27 minutes ago

    Does it make any sense? I tried few models at 128GB and it's all pretty much rubbish. Yes they do give coherent answers, sometimes they are even correct, but most of the time it is just plain wrong. I find it massive waste of time.

    • boutell 18 minutes ago

      I'm not sure how long ago you tried it, but look at Qwen 3.5 32b on a fast machine. Usually best to shut off thinking if you're not doing tool use.

  • S4phyre an hour ago

    Oh how cool. Always wanted to have a tool like this.

  • g_br_l an hour ago

    could you add raspi to the list to see which ridiculously small models it can run?

  • metalliqaz an hour ago

    Hugging Face can already do this for you (with much more up-to-date list of available models). Also LM Studio. However they don't attempt to estimate tok/sec, so that's a cool feature. However I don't really trust those numbers that much because it is not incorporating information about the CPU, etc. True GPU offload isn't often possible on consumer PC hardware. Also there are different quants available that make a big difference.

  • charcircuit an hour ago

    On mobile it does not show the name of the model in favor of the other stats.

  • unfirehose 32 minutes ago

    if you do, would you still want to collect data in a single pane of glass? see my open source repo for aggregating harness data from multiple machine learning model harnesses & models into a single place to discover what you are working on & spending time & money. there is plans for a scrobble feature like last.fm but for agent research & code development & execution.

    https://github.com/russellballestrini/unfirehose-nextjs-logg...

    thanks, I'll check for comments, feel free to fork but if you want to contribute you'll have to find me off of github, I develop privately on my own self hosted gitlab server. good luck & God bless.