Google quietly released an app that lets you download and run AI models locally (github.com)
from Pro@programming.dev to technology@lemmy.world on 31 May 16:03
https://programming.dev/post/31372990

#technology

threaded - newest

rirus@feddit.org on 31 May 16:16 next collapse

Alibaba also provides an OpenSource App, it even has support for their multimodal voice chat Model qwen2.5 omni: github.com/alibaba/MNN

moonlight6205@lemm.ee on 31 May 17:04 next collapse

Is the chat uncensored?

Rhaedas@fedia.io on 31 May 17:26 next collapse

And unmonitored? Don't trust anything from Google anymore.

What makes this better than Ollama?

Angelusz@lemmy.world on 31 May 17:35 next collapse

Quote: “all running locally, without needing an internet connection once the model is loaded. Experiment with different models, chat, ask questions with images, explore prompts, and more!”

So you can download it and set the device to airplane mode, never go online again - they won’t be able to monitor anything, even if there’s code for that included.

taladar@sh.itjust.works on 31 May 17:56 next collapse

That is exactly what Ollama does too.

melroy@kbin.melroy.org on 31 May 19:45 collapse

never go online again - they won't be able to monitor anything, even if there's code for that included.

Sounds counter-intuitive on a smart phone where you most likely want to be online again at some point in time.

Angelusz@lemmy.world on 01 Jun 10:15 collapse

So trust them. If you don’t and want to use this, buy a separate device for it, or VM.

Can’t? This is not for you.

melroy@kbin.melroy.org on 01 Jun 11:16 collapse

I won't gonna use my smartphone as a local llm machine.

GreenKnight23@lemmy.world on 31 May 17:40 collapse

everything is unmonitored if you don’t connect to the network.

flightyhobler@lemmy.world on 31 May 19:21 collapse

But not everything works in those conditions.

GreenKnight23@lemmy.world on 31 May 23:12 collapse

it does if you make it work in those conditions.

software that “phones home” is easy to fool.

Deckname@discuss.tchncs.de on 01 Jun 08:08 collapse

Just firewall the software or is there anything more fancy i would need to do?

GreenKnight23@lemmy.world on 01 Jun 15:46 collapse

typically the phone home is looking for a response to unlock.

use a packet sniffer to see what the request/response is and replicate it with a proxy or response server.

this is also know as a man-in-the-middle (mitm).

takes skill and knowledge to do, but once you do a few dozen it’s pretty easy since most software “phone homes” are looking for static non-encrypted responses.

ofcourse@lemmy.ml on 31 May 18:55 collapse

Censoring is model dependent so you can select one of the models without the guardrails.

AmbiguousProps@lemmy.today on 31 May 17:35 next collapse

Why would I use this over Ollama?

Greg@lemmy.ca on 31 May 18:01 collapse

Ollama can’t run on Android

AmbiguousProps@lemmy.today on 31 May 18:08 next collapse

That’s fair, but I think I’d rather self host an Ollama server and connect to it with an Android client in that case. Much better performance.

Greg@lemmy.ca on 31 May 21:46 next collapse

Yes, that’s my setup. But this will be useful for cases where internet connection is not reliable

OhVenus_Baby@lemmy.ml on 01 Jun 15:22 collapse

How is Ollama compared to GPT models? I used the paid tier for work and I’m curious how this stacks up.

AmbiguousProps@lemmy.today on 01 Jun 17:00 collapse

It’s decent, with the deepseek model anyway. It’s not as fast and has a lower parameter count though. You might just need to try it and see if it fits your needs or not.

Euphoma@lemmy.ml on 31 May 22:15 next collapse

You can use it in termux

Greg@lemmy.ca on 31 May 23:30 collapse

Has this actually been done? If so, I assume it would only be able to use the CPU

Euphoma@lemmy.ml on 01 Jun 04:21 collapse

Yeah I have it in termux. Ollama is in the package repos for termux. The speed it generates does feel like cpu speed but idk

Diplomjodler3@lemmy.world on 01 Jun 11:19 next collapse

Is there any useful model you can run on a phone?

pirat@lemmy.world on 01 Jun 14:54 next collapse

Try PocketPal instead

gens@programming.dev on 01 Jun 15:12 collapse

Llama.cpp (on which ollama runs on) can. And many chat programs for phones can use it.

NGC2346@sh.itjust.works on 31 May 22:13 next collapse

Enclave on iOS does the trick for the rare times i need a local LLM

th3dogcow@lemmy.world on 31 May 23:25 collapse

Didn’t know about this. Checking it out now, thanks!

toastmeister@lemmy.ca on 01 Jun 01:15 next collapse

Duck.ai doesn’t data mine, and has o3 mini which I have found to be very good. Its got some extra functionality like lines to break up text.

stardust@lemmy.ca on 01 Jun 02:05 next collapse

Yeah duck is all over bothered with since it came out since you don’t even need to login to use it.

Kuma@lemmy.world on 01 Jun 08:52 next collapse

Nice! I saw Mozilla also added an ai chat in the browser recently (not in the phone version that I have seen tho).

It is too bad duck.ai only runs the small models. Gpt4o-mini is not very good, it can be very inaccurate and very inconsistent :( I would like to see the 4.1-mini instead, faster and better and got function calling, so it can do web searches for example. O3 can’t so it can only know what it knows until 2023.

But thanks for the information I will be looking out for when 4.1 is added!

Libra@lemmy.ml on 01 Jun 19:25 collapse

I’ve been using duck.ai recently myself and quite like it. My only complaint with it is that the chats have a length limit, so if you’re working on complex projects you can run into those limits pretty quick. I use it for worldbuilding for a novel I’m working on and I have to use chatgpt for thematic stuff because it has a better memory, but otherwise it’s great for quick/small things.

RizzoTheSmall@lemm.ee on 01 Jun 17:51 next collapse

You never heard of ollama or docker model runner?

fmstrat@lemmy.nowsci.com on 02 Jun 01:32 collapse

Android and iOS.

Wazowski@lemmy.world on 01 Jun 21:41 next collapse

Excellent, I will be sure not to use this, like all Google shit.

bizzle@lemmy.world on 02 Jun 00:14 collapse

In a few years you won’t be able to anyway

Kolanaki@pawb.social on 02 Jun 02:16 collapse

I’m just reaching the end game faster then.

rickyrigatoni@lemm.ee on 02 Jun 00:31 collapse

All the time I spent trying to get rid of gemini just to now download this. Am I stupid?

JustARegularNerd@lemmy.dbzer0.com on 02 Jun 02:15 collapse

I wouldn’t think so - it depends on your priorities.

The open source and offline nature of this without the pretenses of “Hey, we’re gonna use every query you give as a data point to shove more products down your face” seems very appealing over Gemini. There’s also that Gemini is constantly being shoved in our faces and preinstalled, whereas this is a completely optional download.