Gemma 3 27B
Google · released Mar 1, 2025 · google/gemma-3-27b-it
- Type
- Open weightsGemma Terms of Use
- Params
- 27.4B
- Context
- 262K
about 197K words of context · download allowed, licence restricts use
Our take
Written Aug 3, 2026Gemma 3 is a 27.4-billion-parameter text-and-image model from Google with broad benchmark coverage and budget-friendly hosted pricing. Its licence carries commercial restrictions, so teams should check the terms before building products around it.
Pick this for budget image-and-text inference when low output cost matters, or for projects needing a 262,144-token request limit at mid-size scale. Use it when breadth of Arena evaluation matters more than top-tier scores. Skip it if you need a permissive open-source licence like Apache or MIT, or if your workload is maths-heavy where its relative performance dips.
The case for it
- Very cheap hosted entry point for multimodal inference, with the cheapest tracked host costing several times less than the most expensive.
- Broad benchmark coverage: six Arena categories measured, with the overall text score stable within a fraction of a point across evaluation dates.
- A strong throughput option exists, at 58 tokens per second on one tracked host.
The case against it
- Gemma Terms of Use restrict commercial freedom compared with Apache or MIT alternatives; redistribution and commercial use are subject to Google-specific terms.
- No disclosed active parameter count, so the true inference cost per token is obscured.
- Maths performance lags its own other Arena categories by the widest spread in its benchmark scores.
How good is it?
IntelligencePuzzles, maths, exam questions
Arena Text (overall)94th of 143 · 1365.9
CodingWriting and fixing code on its own
Arena Coding116th of 143 · 1357.9
Arena Coding is the only board that has scored it for this.
AgenticPlanning, calling tools, staying on task
Nobody we watch has scored Gemma 3 27B 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 3 27B 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. 3.6 GB spare means a 10% error in the size would not change the answer.
GeForce RTX 5090 · 32 GB
Room to spare. 11.6 GB spare means a 10% error in the size would not change the answer.
Apple M1 Pro (16-core GPU) · 32 GB
Room to spare. 4.8 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 8 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
- $0.080 in / $0.45 out
- Context served
- 262K
- Throughput
- Not measured
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| DeepInfrafp8 | $0.080 / $0.16 | 131K | not measured | Unknown | Unknown | Unknown |
| DeepInfrafp8 | $0.080 / $0.16 | 131K | 24 tok/s | No | No | Confirmed |
| Novita AIbf16 | $0.12 / $0.20 | 98K | 35 tok/s | No | No | Confirmed |
| Novita AI | $0.12 / $0.20 | 98K | not measured | Unknown | Unknown | Unknown |
| Nebius AI Studiofp8 | $0.10 / $0.30 | 110K | 38 tok/s | No | No | Confirmed |
| Parasailfp8 | $0.080 / $0.45 | 131K | 50 tok/s | No | No | Confirmed |
| OpenRouter | $0.080 / $0.45 | 262K | not measured | Unknown | Unknown | Unknown |
| Phala | $0.15 / $0.46 | 262K | 15 tok/s | No | No | Confirmed |
DeepInfra
Novita AI
Nebius AI StudioAcross the 8 listings we hold: 5 say they do not train on prompts, 0 say they do and 3 do not say. 5 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 |
|---|---|---|---|
| DeepInfrafp8 | |||
| DeepInfrafp8 | ✓ | ✓ | ✓ |
| Novita AIbf16 | ✗ | ✗ | ✗ |
| Novita AI | |||
| Nebius AI Studiofp8 | ✗ | ✓ | ✓ |
| Parasailfp8 | ✗ | ✓ | ✓ |
| OpenRouter | ✓ | ✓ | ✓ |
| Phala | ✗ | ✓ | ✓ |
Tool calling: 2 of 8 listings say yes, 4 say no, 2 publish no parameter list. JSON output: 5 of 8 listings say yes, 1 says no, 2 publish no parameter list. Strict schema: 5 of 8 listings say yes, 1 says no, 2 publish no parameter list.
Models people weigh against Gemma 3 27B
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.
- 2 of 8 listings publish no parameter list, so what their API accepts is unknown to us.
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- 3 of 8 listings do not say whether they train on prompts.
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-3-27b-it
- Modality record
- text+image->text
- Catalogue slug
- google-gemma-3-27b