from Hacker News

Google AI Edge – On-device cross-platform AI deployment

by nreece on 6/1/25, 6:32 AM with 43 comments

• by salamo on 6/1/25, 11:04 PM

  Really happy to see additional solutions for on-device ML.

  That said, I probably wouldn't use this unless mine was one of the specific use cases supported[0]. I have no idea how hard it would be to add a new model supporting arbitrary inputs and outputs.

  For running inference cross-device I have used Onnx, which is low-level enough to support whatever weights I need. For a good number of tasks you can also use transformers.js which wraps onnx and handles things like decoding (unless you really enjoy implementing beam search on your own). I believe an equivalent link to the above would be [1] which is just much more comprehensive.

  [0] https://ai.google.dev/edge/mediapipe/solutions/guide

  [1] https://github.com/huggingface/transformers.js-examples

• by arbayi on 6/1/25, 3:44 PM

  https://github.com/google-ai-edge/gallery

  A gallery that showcases on-device ML/GenAI use cases and allows people to try and use models locally.

• by ricardobeat on 6/1/25, 1:35 PM

  This is a repackaging of TensorFlow Lite + MediaPipe under a new “brand”.

• by pzo on 6/2/25, 4:29 AM

  My take: tensorflow lite + mediapipe was great but google really neglected it in the last 3 years or so. Mediapipe didn't have many meaningful update in last 3 years. A lot of models today are outdated or slow. TF Lite supported NPU (like apple ANU) but mediapipe never did. They had also too much mess with different branding: MLKit, Firebase ML, TF lite, LiteRT.

  This days probably better to stick with onnxruntime via hugging face transformers or transformers.js library or wait until executorch mature. I haven't seen any SOTA model officially released having official port to tensorflow lite / liteRT for a long time: SAM2, EfficientSAM, EdgeSAM, DFINE, DEIM, Whisper, Lite-Whisper, Kokoro, DepthAnythingV2 - everything is pytorch by default but with still big communities for ONNX and MLX

• by yeldarb on 6/1/25, 12:53 PM

  Is this a new product or a marketing page tying together a bunch of the existing MediaPipe stuff into a narrative?

  Got really excited then realized I couldn’t figure out what “Google AI Edge” actually _is_.

  Edit: I think it’s largely a rebrand of this from a couple years ago: https://developers.googleblog.com/en/introducing-mediapipe-s...

• by 6gvONxR4sf7o on 6/2/25, 12:47 AM

  Anybody have any experience with this? I just spend a while contorting a custom pytorch model to get it to export to coreml and it was full of this that and the other not being supported, or segfaulting, and all sorts of silly errors. I'd love if someone could say this isn't full of sharp edges too.

• by davedx on 6/1/25, 12:25 PM

  More information here: https://ai.google.dev/edge/mediapipe/solutions/guide

  (It seems to be open source: https://github.com/google-ai-edge/mediapipe)

  I think this is a unified way of deploying AI models that actually run on-device ("edge"). I guess a sort of "JavaScript of AI stacks"? I wonder who the target audience is for this technology?

• by hatmanstack on 6/1/25, 3:13 PM

  Played with this a bit and from what I gathered it's purely a re-arch of pytorch models to work as .tflite models, at least that's what I was using it for. It worked well with a custom finbert model with negligible size reduction. It converted a quantized version but outputs were not close. From what I remember of the docs it was created for standard pytorch models, like "torchvision.models", so maybe with those you'd have better luck. Granted, this was all ~12 months ago, sounds like I might have dodged a pack of Raptors?

• by zb3 on 6/1/25, 1:24 PM

  So can we run Gemma 3n on linux? So much fluff yet this is unclear to me.

• by danielb123 on 6/1/25, 1:22 PM

  Years behind what is already available through frameworks like CoreML and TimyML. Plus Google has to first prove they won't kill the product to meet the next quarterly investor expectations.

• by rs186 on 6/3/25, 12:18 AM

  LOL when you realize that Google wants you to download the apk and sideload it instead of installing from the Play Store ( https://github.com/google-ai-edge/gallery#-get-started-in-mi... )

  You know how terrible the store and the publishing process are -- their own people don't even use it.

• by dingody on 6/3/25, 2:30 AM

  I keep seeing people talk a lot about edge AI — just curious, aside from those experimental or toy projects, are there any real killer use cases out there?

• by stanleykm on 6/1/25, 3:40 PM

  i really wish people who make edge inference libraries like this would quit rebranding them every year and just build the damn things to be fast and small and consistently updated.

• by roflcopter69 on 6/1/25, 10:52 PM

  Genuine question, why should I use this to deploy models on the edge instead of executorch? https://github.com/pytorch/executorch

  For context, I get to choose the tech stack for a greenfield project. I think that executor h, which belongs to the pytorch ecosystem, will have a way more predictable future than anything Google does, so I currently consider executorch more.

• by init0 on 6/2/25, 4:06 AM

  This can be done with WebLLM, no?

• by synergy20 on 6/2/25, 1:45 AM

  Can this run on customized embedded devices? or just for phones.

• by suilk on 6/2/25, 1:36 AM

  How about the MNN engine?

• by rvnx on 6/1/25, 1:07 PM

  Make your own opinion here: https://mediapipe-studio.webapps.google.com/studio/demo/imag...

  Go to this page using your mobile phone.

  I am apparently a doormat or a seatbelt.

  It seems to be a rebranded failure. At Google you get promoted for product launches because of the OKRs system and more rarely for maintenance.