Gemma 3n 4B
Google · released Jun 3, 2025 · google/gemma-3n-E4B-it
- Type
- Open weightsGemma Terms of Use
- Params
- 7.8B
- Context
- 33K
about 25K words of context · download allowed, licence restricts use
Our take
Written Aug 3, 2026Gemma 3n is a small downloadable text model from Google with 7.8 billion parameters and a 32,768-token request limit. Its scores on the chat leaderboard have barely moved between recent evaluation runs, suggesting consistent measured performance for a budget-priced offering.
Pick this for low-cost hosted text generation where you want identical pricing across providers and no need to hunt for arbitrage. Use it for local deployment on modest hardware without the complexity of a mixture-of-experts architecture. Skip it if you need image or video input, a longer request limit, or a fully permissive licence without redistribution restrictions.
The case for it
- Identical pricing across both tracked hosts, so there is no markup arbitrage to chase.
- Arena Text Elo moved less than a tenth of a point between July and August 2026 evaluation runs, suggesting reproducible measured performance.
- Known throughput of 26 tokens per second on one disclosed host.
The case against it
- Text-only with a 32,768-token cap, while competitors often add vision or longer context.
- Gemma Terms of Use carry redistribution restrictions, unlike an Apache licence.
- Maths is a clear weak spot: its Arena Maths score sits almost 59 points below its own overall text score.
How good is it?
IntelligencePuzzles, maths, exam questions
Arena Text (overall)117th of 143 · 1318.2
CodingWriting and fixing code on its own
Arena Coding130th of 143 · 1307.7
Arena Coding is the only board that has scored it for this.
AgenticPlanning, calling tools, staying on task
Nobody we watch has scored Gemma 3n 4B for this. We would take the rating from Arena Agent (IPS).
WritingWe do not rate this
Two boards come close and neither tests writing: Arena Creative Writing asks people which of two replies they prefer, and LiveBench Language tests whether a model understood a passage. So we show where Gemma 3n 4B placed and give it no mark out of five.
These tests check whether a model follows instructions — a precondition for all the work above, but not a measure of how well that work is done, which is why they get no rating.
Every published score for this model6 scoresEvery figure we hold, from 6 boards, with who ran it and a link to the source — including the boards no rating above is built on.
Can you run it yourself?
- Fits in memory
- weights load entirely on the card
- Spills to system RAM
- some weights offload; much slower
- Too large
- will not load even with offload
- est
- size is calculated; the verdict could change by 10%
Comfortable fit
GeForce RTX 4090 · 24 GB
Room to spare. 16.4 GB spare means a 10% error in the size would not change the answer.
GeForce RTX 5090 · 32 GB
Room to spare. 24.4 GB spare means a 10% error in the size would not change the answer.
Apple M1 (8-core GPU) · 16 GB
Room to spare. 5.6 GB spare means a 10% error in the size would not change the answer.
Memory use by level
Against a 24 GB card.
All 3 sizes here are calculated, not measured. We hold no measured file for this model, so each size comes from the parameter count and every verdict above inherits that uncertainty.
Check against your own machine → · All 71 devices, with every size →
Or rent it from someone else
Cheapest of 2 live listings. Picked at the widest standard context we hold, within one quantisation slice, so the numbers beside it are a price one host actually charges.
- per 1M tokens
- $0.060 in / $0.12 out
- Context served
- 33K
- Throughput
- Not measured
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| OpenRouter | $0.060 / $0.12 | 33K | not measured | Unknown | Unknown | Unknown |
| Together AI | $0.060 / $0.12 | 33K | 32 tok/s | No | No | Confirmed |
Together AIAcross the 2 listings we hold: 1 say they do not train on prompts, 0 say they do and 1 do not say. 1 appear in the zero-retention registry we check; the rest are unknown to us rather than confirmed either way.
What each host's API supports
From the parameter list each endpoint publishes. Streaming is omitted: nothing we hold reports it, for any model.
- ✓
- Supported
- ✗
- Not supported
- Not published
- host gave no parameter list
| Provider | Tool calling | JSON output | Strict schema |
|---|---|---|---|
| OpenRouter | ✗ | ✓ | ✓ |
| Together AI | ✗ | ✓ | ✓ |
Tool calling: 0 of 2 listings say yes, 2 say no. JSON output: 2 of 2 listings say yes. Strict schema: 2 of 2 listings say yes.
Models people weigh against Gemma 3n 4B
When we formed this view
Dates behind this page
Prices last checked 6h ago
What we do not know about this model yet
- We hold no measured file for it, so all 3 sizes on this page are calculated from the parameter count.
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- 1 of 2 listings do not say whether they train on prompts.
- We hold no cached-input rate for any of its listings.
Licence and identifiers
What the licence allowsGemma Terms of Use, what it allows commercially, and the identifiers you need to pull this model — its Hugging Face repo, our slug and a machine-readable card.
Licence
Gemma Terms of Use
Commercial use allowed, but Google's prohibited-use policy applies and can be updated over time — terms are less static than Apache/MIT.
Identifiers
- Hugging Face
- google/gemma-3n-E4B-it
- Modality record
- text->text
- Catalogue slug
- google-gemma-3n-4b