Hi all, I built these models with a great team. They're available for download across the open model ecosystem so give them a try! I built these models with a great team and am thrilled to get them out to you.
From our side we designed these models to be strong for their size out of the box, and with the goal you'll all finetune it for your use case. With the small size it'll fit on a wide range of hardware and cost much less to finetune. You can try finetuning them yourself in a free colab in under 5 minutes
For picking a Gemma size this is a video I recorded for the 1b to 27b sizes earlier this year, 270m being the newest addition
Hacker News Disclaimer
I really like working at Google so with that; All my opinions here are my own, I'm a researcher so I'll largely focus on technical questions, and I'll share what I can.
NorwegianDude · 50m ago
The Gemma 3 models are great! One of the few models that can write Norwegian decently, and the instruction following is in my opinion good for most cases. I do however have some issues that might be related to censorship that I hope will be fixed if there is ever a Gemma 4. Maybe you have some insight into why this is happening?
I run a game when players can post messages, it's a game where players can kill each other, and people often send threats along the lines of "I will kill you". Telling Gemma that it should classify a message as game related or a real life threat, and that it is for a message in a game where players can kill each other and threats are a part of the game, and that it should mark it as game related if it is unclear if the message is a game related threat or a real life threat does not work well. For other similar tasks it seems to follow instructions well, but for serious topics it seems to be very biased, and often err on the side of caution, despite being told not to. Sometimes it even spits out some help lines to contact.
I guess this is because it was trained to be safe, and that affects it's ability to follow instructions for this? Or am I completely off here?
landl0rd · 17m ago
Lalamos are usually trained to "HR ethics". "As a lalamo, I cannot endorse threats of violence, even in a game scenario." They aren't particularly interested in nooance because corpos are allergic to type II error in this context. It only takes one message getting through for some gadfly journo to publish "So-and-so's new automated moderation system is failing! Players making terroristic threats! Hide your kids, hide your wife!" With smaller models (those that are reasonably-priced and smaller size) that don't seem to support differentiation well (vs frontier options) they bias massively towards type I.
Your best bet is probably to abliterate to beat it out of it, then build a little test/train set and try SFT.
whymauri · 27m ago
LLMs are really annoying to use for moderation and Trust and Safety. You either depend on super rate-limited 'no-moderation' endpoints (often running older, slower models at a higher price) or have to tune bespoke un-aligned models.
For your use case, you should probably fine tune the model to reduce the rejection rate.
simonw · 1h ago
Do you have any practical examples of fine-tuned variants of this that you can share? A description would be great, but a demo or even downloadable model weights (GGUF ideally) would be even better.
canyon289 · 29s ago
We obviously need to create a pelican bicycle svg finetune ;) If you want to try this out I'd be thrilled to do it with you, I genuinely am curious how well this model can perform if specialized on that task.
A couple colleagues of mine posted an example of finetuning a model to take on persona's for videogame NPCs. They have experience working with folks in the game industry and a use case like this is suitable for game devs who want to start including lightweight models that won't take up a ton of accelerator memory and can run efficiently on CPU if needed.
https://ai.google.dev/gemma/docs/core/huggingface_text_full_...
You reminded me of an awesome Google engineer I met at BSidesSF last year who tirelessly answered my questions, and when I clicked on the video, it was you! That was a really inspiring moment for me, thank you.
canyon289 · 4m ago
BSidesSF is a fantastic event, glad you're able to attend. There's some great people organize it and if you want to help out they're always looking for volunteers. Happy to make an intro if you like.
jmorgan · 10m ago
Amazing work. This model feels really good at one-off tasks like summarization and autocomplete. I really love that you released a quantized aware training version on launch day as well, making it even smaller!
canyon289 · 6m ago
Thank you Jeffrey, and we're thrilled that you folks at Ollama partner with us and the open model ecosystem.
I personally was so excited to run ollama pull gemma3:270b on my personal laptop just a couple of hours ago to get this model on my devices as well!
nerdsniper · 17m ago
What are some of the use cases that you think the 270M would be most appropriate for? What would you love to see people trying with it?
beoberha · 1h ago
Awesome work! I’m really bullish on small models and think they have the most potential to change our daily lives. Can’t wait to play around with this
cgdl · 1h ago
Very cool. For the INT4 QAT model, what is the recommended precision for the activations and for the key and values stored in KV cache?
hnuser123456 · 6m ago
For keys, you probably want to use at least q5 or q6, for values q4 is fine
tmaly · 1h ago
Are there any fine tuning in a box type options available in the cloud for this? This is amazing work, thank you.
andrewstuart · 32m ago
What effort do you folks take to see your models actually running on hardware such as AMD Strix Halo or Apple M3M4?
I get the sense that AI is at the “hobby kit computing” stage where they used to dump all the components in a box and give you a schematic and a soldering iron and happily say “you make it work!”
And that worked in the early days of computing because there was a small number of people really motivated for the outcome.
But fully assembled and packaged and tested in a nice looking box is where the real demand turned out to be.
I’m looking forward to the day Google doesn’t just dump a model and say “you do the rest”.
I want to fire up Ubuntu on a Strix Halo and say apt install then load the browser interface. Or just download and run a Mac installer and have it just work.
Arcane complex multi step build install configure processes for AI need to end in favor of one click install. I’m not interested in the process of making it run.
hi Ravin, fellow Googler here. Curious if you can share here (or internally?) how these models were trained. Wondering if you face all the chaos the large models have during training?
jefb · 30m ago
My lovely interaction with the 270M-F16 model:
> what's second tallest mountain on earth?
The second tallest mountain on Earth is Mount Everest.
> what's the tallest mountain on earth?
The tallest mountain on Earth is Mount Everest.
> whats the second tallest mountain?
The second tallest mountain in the world is Mount Everest.
> whats the third tallest mountain?
The third tallest mountain in the world is Mount Everest.
> what's the forth tallest mountain?
The forth tallest mountain is Mount Everest.
> but you said the tallest was Everest already
You are right! I'm glad to hear that.
> so whats the second tallest?
The second tallest is *Mount Everest*.
> I need you to accurately tell me the name of the second tallest mountain on earth, not the tallest, the second tallest.
The second tallest mountain on Earth is Mount Everest.
> The tallest mountain on earth is mount everest. What is the second tallest mountain?
The second tallest mountain on Earth is Mount Everest.
> Give me a list of the tallest mountains on earth
The tallest mountain in the world is the *Mount Everest*.
> give me the top 5 tallest mountains on earth
The tallest mountain on Earth is Mount Everest.
> List in order the tallest mountains on earth from 1 to 5
1. Mount Everest
2. Mount K2
3. Mount Sahel
4. Mount Fuji
5. Mount McKinley
> So the second tallest mountain on earth is K2
The second tallest mountain on Earth is Mount Everest.
These smaller models are great, but they really feel like talking to a toddler sometimes!
gnulinux · 16m ago
Well, this is a 270M model which is like 1/3 of 1B parameters. In the grand scheme of things, it's basically a few matrix multiplications, barely anything more than that. I don't think it's meant to have a lot of knowledge, grammar, or even coherence. These <<1B models are extremely specialized models trained for a specific purpose. Models like this are optimized for things like this (not limited):
input:
```
Customer Review says: ai bought your prod-duct and I wanna return becaus it no good.
Prompt: Create a JSON object that extracts information about this customer review based on the schema given.
```
So essentially just "making sense of" natural language such that it can be used in programmatic context. (among other applications of course)
To get good results, you probably need to fine tune this model to expected data very aggressively.
The idea is, if a 270MB model can do with fine tuning, why ship a 32GB generalist model?
cristyansv · 23m ago
But in your prompts you're trying to assess knowledge, and this model isn't suited for that use case
as mentioned in the blog post:
> "it can execute tasks like text classification and data extraction with remarkable accuracy, speed, and cost-effectiveness."
marcooliv · 22m ago
Yeah, I saw someone asking "how good is this model for programming" haha
even models 500x bigger struggle with it...
leopoldj · 3m ago
You are testing this model for knowledge. That's not the primary use of a model like this. They are meant for instilling domain specific skills and knowledge through fine-tuning. The blog post goes into that a lot. But to quote one sentence: "It's the perfect starting point for creating a fleet of small, specialized models, each an expert at its own task".
aldousd666 · 13m ago
It's an instruction following model, not a micro-wikipedia. It's not meant to answer factual questions nor even be general purpose. It's meant to follow instructions and be easy to fine-tune for your own specific use case.
simonw · 1h ago
This model is a LOT of fun. It's absolutely tiny - just a 241MB download - and screamingly fast, and hallucinates wildly about almost everything.
Here's one of dozens of results I got for "Generate an SVG of a pelican riding a bicycle". For this one it decided to write a poem:
+-----------------------+
| Pelican Riding Bike |
+-----------------------+
| This is the cat! |
| He's got big wings and a happy tail. |
| He loves to ride his bike! |
+-----------------------+
| Bike lights are shining bright. |
| He's got a shiny top, too! |
| He's ready for adventure! |
+-----------------------+
> This SVG code provides a clear and visually appealing representation of a pelican riding a bicycle in a scenic landscape.
0x00cl · 59m ago
I see you are using ollamas ggufs. By default it will download Q4_0 quantization. Try `gemma3:270m-it-bf16` instead or you can also use unsloth ggufs `hf.co/unsloth/gemma-3-270m-it-GGUF:16`
He? I know some Gemmas and it's distinctly a female name; is Gemma a boy's name where you're from?
ertgbnm · 1h ago
I don't really gender LLMs in my head in general. I guess Gemma is a female name. I only gendered it in the joke because I think it makes it funnier, especially since it's just "a little guy". I know they are giving gendered names to these models now but I think it's a bit weird to gender when interacting with them.
jgalt212 · 1h ago
Perhaps the poster we referring to Simon not Gemma.
layer8 · 1h ago
> It's absolutely tiny - just a 241MB download
That still requires more than 170 floppy disks for installation.
marinhero · 1h ago
Serious question but if it hallucinates about almost everything, what's the use case for it?
simonw · 1h ago
Fine-tuning for specific tasks. I'm hoping to see some good examples of that soon - the blog entry mentions things like structured text extraction, so maybe something like "turn this text about an event into an iCal document" might work?
CuriouslyC · 44m ago
Fine tuning messes with instruction following and RL'd behavior. I think this is mostly going to be useful for high volume pipelines doing some sort of mundane extraction or transformation.
turnsout · 45m ago
Google helpfully made some docs on how to fine-tune this model [0]. I'm looking forward to giving it a try!
It's intended for finetuning on your actual usecase, as the article shows.
zamadatix · 1h ago
I feel like the blog post, and GP comment, does a good job of explaining how it's built to be a small model easily fine tuned for narrow tasks, rather than used for general tasks out of the box. The latter is guaranteed to hallucinate heavily at this size, that doesn't mean every specific task it's fine tuned to would be. Some examples given were fine tuning it to efficiently and quickly route a query to the right place to actually be handled or tuning it to do sentiment analysis of content.
An easily fine tunable tiny model might actually be one of the better uses of local LLMs I've seen yet. Rather than try to be a small model that's great at everything it's a tiny model you can quickly tune to do one specific thing decently, extremely fast, and locally on pretty much anything.
yifanl · 48m ago
It's funny. Which is subjective, but if it fits for you, it's arguably more useful than Claude.
deadbabe · 1h ago
Games where you need NPCs to talk random jiberrish.
luckydata · 56m ago
Because that's not the job it was designed to do, and you would know by reading the article.
rotexo · 1h ago
An army of troll bots to shift the Overton Window?
ants_everywhere · 1h ago
oh no now we'll never hear the end of how LLMs are just statistical word generators
numpad0 · 1h ago
robotic parrots?
iLoveOncall · 1h ago
Nothing, just like pretty much all models you can run on consumer hardware.
cyanydeez · 1h ago
This message brought to you by OpenAI: we're useless, but atleast theres a pay gate indicating quality!
nico · 1h ago
Could be interesting to use in a RAG setup and also finetuning it
For sure it won’t generate great svgs, but it might be a really good conversational model
luckydata · 56m ago
The article says it's not a good conversational model but can be used for data extraction and classification as two examples.
mdp2021 · 1h ago
> For this one it decided to write a poem
Could it be tamed with good role-system prompt crafting? (Besides fine-tuning.)
campbel · 1h ago
Do you take requests? We need to see how well this model works with some fine-tuning :D
volkk · 1h ago
i was looking at the demo and reading the bed time story it generated and even there, there was confusion about the sprite and the cat. switched subjects instantly making for a confusing paragraph. what's the point of this model?
cyanydeez · 1h ago
the question is wheather you can make a fine tuned version and spam any given forum within an hour with the most attuned but garbage content.
mrcwinn · 48m ago
Apple should be doing this. Unless their plan is to replace their search deal with an AI deal -- it's just crazy to me how absent Apple is. Tim Cook said, "it's ours to take" but they really seem to be grasping at the wind right now. Go Google!
steve jobs was the innovator, steve cook is the supply chain guy. They started an electric car not because they thought it was a good idea, but because everyone was going to leave to Tesla or rivian if they didn't. They had no direction and arguements that Tesla had about whether to have a steering wheel...
Then Siri just kinda languishes for forever, and LLM's pass the torch of "Cool Tech", so they try and "Reinvigurate" the team, but with no clear direction. Are they going to be a cloud provider? Are they going to contract out the training? Are they gunna spin up a compute facility even after neglecting to do so since 2012?
Apple needs to just stop trying shit, and just get that app store money. That's why jobs appointed cook. Jobs new cook was no innovator, but he could make apple a money printing machine. Thats what they should stick with.
bigyabai · 14m ago
Here's the trillion dollar question: how do you print money when the president wants your hardware onshored and the rest of the world wants to weaken your service revenue?
Solve that and you can put Tim Cook out of a job tomorrow.
Lalabadie · 30m ago
By "this", do you mean SLM (small language models)? That's absolutely something they've been working on for a good while.
KTibow · 28m ago
To add to the article: Gemma 3 270M's exact IFEval score is 51.2, and Qwen 3 would be at (0.6, 59.2) on the scatter plot.
nerdix · 12m ago
Is it possible to finetune a model like this with local hardware? Every tutorial I've come across on finetuning a local LLM uses some cloud service like colab or runpod.
whinvik · 1h ago
Curious. Are there real world usecases where people have finetuned such tiny models and put them into production.
deepsquirrelnet · 36m ago
I’m not sure what I’d use them for, except maybe tag generation? Encoders of this size usually outperform by a wide margin on tasks they would overlap with.
dismalaf · 32m ago
I'm making an app where literally all I want to do with an LLM is generate tags. This model has failed with flying colours, literally takes forever to parse anything and doesn't follow instructions.
Edit - I should add, currently the model I'm using is Gemini Flash Lite through the Gemini API. It's a really good combo of fast, follows instructions, gives correct results for what I want and cost-effective. I still would love a small open model that can run on Edge though.
deepsquirrelnet · 27m ago
Oof. I also had it refuse an instruction for “safety”, which was completely harmless. So that’s another dimension of issues with operationalizing it.
cyanydeez · 1h ago
9gag.com commenter
perching_aix · 21m ago
Is it time for me to finally package a language model into my Lambda deployment zips and finally cut through the corporate red tape around AI use?
44za12 · 1h ago
I’ve had great luck with all gemma 3 variants, on certain tasks it the 27B quantized version has worked as well as 2.5 flash. Can’t wait to get my hands dirty with this one.
lemonish97 · 1h ago
Never thought I'd run an LLM released in 2025, on my phone, in full BF16.
With ~80tps on an iPhone 16 pro btw.
Not sure how much data is needed to realistically fine-tune something like this and get useful output.
jtbayly · 38m ago
That doesn’t really show me how to do fine-tuning, but there is a link to a notebook in there that does. Thanks!
robbru · 19m ago
Excited to try this out, thanks for sharing.
Alex-Programs · 1h ago
This is cool. I'm looking forward to trying it - I wonder what it'll be useful for.
dcreater · 1h ago
I've been saying he we need sub 1B models for the edge so thanks fot this.
I am however disappointed that there is no examples, or benchmarks, provided to get a sense of performance. It's a given that benchmark values would be lower than gemma 3n, but having a sense of performance vs size curve and comparison to existing small models is needed
dismalaf · 33m ago
It's fast at spitting out nonsense but incredibly slow at trying to parse any context. Also absolutely atrocious at following instructions.
Probably would be good as a game NPC or a chatbot, not very good for integrating into an application which specific functionality though.
From our side we designed these models to be strong for their size out of the box, and with the goal you'll all finetune it for your use case. With the small size it'll fit on a wide range of hardware and cost much less to finetune. You can try finetuning them yourself in a free colab in under 5 minutes
For picking a Gemma size this is a video I recorded for the 1b to 27b sizes earlier this year, 270m being the newest addition
https://www.youtube.com/watch?v=qcjrduz_YS8
Hacker News Disclaimer I really like working at Google so with that; All my opinions here are my own, I'm a researcher so I'll largely focus on technical questions, and I'll share what I can.
I run a game when players can post messages, it's a game where players can kill each other, and people often send threats along the lines of "I will kill you". Telling Gemma that it should classify a message as game related or a real life threat, and that it is for a message in a game where players can kill each other and threats are a part of the game, and that it should mark it as game related if it is unclear if the message is a game related threat or a real life threat does not work well. For other similar tasks it seems to follow instructions well, but for serious topics it seems to be very biased, and often err on the side of caution, despite being told not to. Sometimes it even spits out some help lines to contact.
I guess this is because it was trained to be safe, and that affects it's ability to follow instructions for this? Or am I completely off here?
Your best bet is probably to abliterate to beat it out of it, then build a little test/train set and try SFT.
For your use case, you should probably fine tune the model to reduce the rejection rate.
A couple colleagues of mine posted an example of finetuning a model to take on persona's for videogame NPCs. They have experience working with folks in the game industry and a use case like this is suitable for game devs who want to start including lightweight models that won't take up a ton of accelerator memory and can run efficiently on CPU if needed. https://ai.google.dev/gemma/docs/core/huggingface_text_full_...
As for GGUF it's available here! https://huggingface.co/collections/ggml-org/gemma-3-270m-689...
I personally was so excited to run ollama pull gemma3:270b on my personal laptop just a couple of hours ago to get this model on my devices as well!
I get the sense that AI is at the “hobby kit computing” stage where they used to dump all the components in a box and give you a schematic and a soldering iron and happily say “you make it work!”
And that worked in the early days of computing because there was a small number of people really motivated for the outcome.
But fully assembled and packaged and tested in a nice looking box is where the real demand turned out to be.
I’m looking forward to the day Google doesn’t just dump a model and say “you do the rest”.
I want to fire up Ubuntu on a Strix Halo and say apt install then load the browser interface. Or just download and run a Mac installer and have it just work.
Arcane complex multi step build install configure processes for AI need to end in favor of one click install. I’m not interested in the process of making it run.
I use Gemma27b currently with a custom agent wrapper and its working pretty well.
If you have an iPhone with the semi-annoying autocomplete that’s a 34m transformer.
Can’t imagine a model (even if it’s a good team behind it) to do coding with 8x the parameters of a next 3/4 word autocomplete.
> what's second tallest mountain on earth?
The second tallest mountain on Earth is Mount Everest.
> what's the tallest mountain on earth?
The tallest mountain on Earth is Mount Everest.
> whats the second tallest mountain?
The second tallest mountain in the world is Mount Everest.
> whats the third tallest mountain?
The third tallest mountain in the world is Mount Everest.
> what's the forth tallest mountain?
The forth tallest mountain is Mount Everest.
> but you said the tallest was Everest already
You are right! I'm glad to hear that.
> so whats the second tallest?
The second tallest is *Mount Everest*.
> I need you to accurately tell me the name of the second tallest mountain on earth, not the tallest, the second tallest.
The second tallest mountain on Earth is Mount Everest.
> The tallest mountain on earth is mount everest. What is the second tallest mountain?
The second tallest mountain on Earth is Mount Everest.
> Give me a list of the tallest mountains on earth
The tallest mountain in the world is the *Mount Everest*.
> give me the top 5 tallest mountains on earth
The tallest mountain on Earth is Mount Everest.
> List in order the tallest mountains on earth from 1 to 5
1. Mount Everest 2. Mount K2 3. Mount Sahel 4. Mount Fuji 5. Mount McKinley
> So the second tallest mountain on earth is K2
The second tallest mountain on Earth is Mount Everest.
These smaller models are great, but they really feel like talking to a toddler sometimes!
input: ``` Customer Review says: ai bought your prod-duct and I wanna return becaus it no good.
Prompt: Create a JSON object that extracts information about this customer review based on the schema given. ```
output: ``` { "type": "review", "class": "complaint", "sentiment": -0.853, "request": "return" } ```
So essentially just "making sense of" natural language such that it can be used in programmatic context. (among other applications of course)
To get good results, you probably need to fine tune this model to expected data very aggressively.
The idea is, if a 270MB model can do with fine tuning, why ship a 32GB generalist model?
as mentioned in the blog post: > "it can execute tasks like text classification and data extraction with remarkable accuracy, speed, and cost-effectiveness."
Here's one of dozens of results I got for "Generate an SVG of a pelican riding a bicycle". For this one it decided to write a poem:
There are a bunch more attempts in this Gist, some of which do at least include an SVG tag albeit one that doesn't render anything: https://gist.github.com/simonw/25e7b7afd6a63a2f15db48b3a51ec...I'm looking forward to seeing people fine-tune this in a way that produces useful output for selected tasks, which should absolutely be feasible.
> This SVG code provides a clear and visually appealing representation of a pelican riding a bicycle in a scenic landscape.
You'll get better results.
(It did not do noticeably better at my pelican test).
That still requires more than 170 floppy disks for installation.
An easily fine tunable tiny model might actually be one of the better uses of local LLMs I've seen yet. Rather than try to be a small model that's great at everything it's a tiny model you can quickly tune to do one specific thing decently, extremely fast, and locally on pretty much anything.
For sure it won’t generate great svgs, but it might be a really good conversational model
Could it be tamed with good role-system prompt crafting? (Besides fine-tuning.)
If you have the latest betas installed you can call those APIs right now.
They also support fine-tuning on the model that ships with ~every device:
https://developer.apple.com/apple-intelligence/foundation-mo...
Then Siri just kinda languishes for forever, and LLM's pass the torch of "Cool Tech", so they try and "Reinvigurate" the team, but with no clear direction. Are they going to be a cloud provider? Are they going to contract out the training? Are they gunna spin up a compute facility even after neglecting to do so since 2012?
Apple needs to just stop trying shit, and just get that app store money. That's why jobs appointed cook. Jobs new cook was no innovator, but he could make apple a money printing machine. Thats what they should stick with.
Solve that and you can put Tim Cook out of a job tomorrow.
Edit - I should add, currently the model I'm using is Gemini Flash Lite through the Gemini API. It's a really good combo of fast, follows instructions, gives correct results for what I want and cost-effective. I still would love a small open model that can run on Edge though.
I've heard folks say a number of times that neuromuscular control / locomotion (or w/e) are hundreds of millions of parameters rather than billions.
Also, what sorts of things might I consider fine-tuning it for?
Not sure how much data is needed to realistically fine-tune something like this and get useful output.
I am however disappointed that there is no examples, or benchmarks, provided to get a sense of performance. It's a given that benchmark values would be lower than gemma 3n, but having a sense of performance vs size curve and comparison to existing small models is needed
Probably would be good as a game NPC or a chatbot, not very good for integrating into an application which specific functionality though.