Gemma 4 26B A4B
Google · released Mar 11, 2026 · google/gemma-4-26B-A4B-it
- Type
- Open weightsApache License 2.0
- Params
- 26.5B
- Context
- 262K
4B active per word · about 197K words of context
Our take
Written Aug 2, 2026Google's Gemma is a mid-size downloadable language model with a permissive Apache licence. It uses an efficient mixture-of-experts design that activates only a few billion parameters per token, making it the efficiency pick of the mid-size class.
Make this your default local mid-size pick: measured chat quality, a permissive licence and efficient design. Use it for cheap hosted multimodal inference in our catalogue, or for edge deployment via Cloudflare Workers AI. Skip it if you need top-tier chat quality or the absolute lowest hosting count.
The case for it
- Measured chat quality unusually close to models many times larger.
- Apache 2.0 licence allows commercial use, fine-tuning and redistribution.
- Only about four billion active parameters per token from a 26.5-billion total, making it efficient for local use.
The case against it
- Trails the frontier on measured chat quality.
How good is it?
IntelligencePuzzles, maths, exam questions
Arena Text (overall)47th of 143 · 1438.5
CodingWriting and fixing code on its own
Arena Coding50th of 143 · 1480.7
AgenticPlanning, calling tools, staying on task
Nobody we watch has scored Gemma 4 26B A4B 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 4 26B A4B 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 model7 scoresEvery figure we hold, from 7 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. 3.9 GB spare means a 10% error in the size would not change the answer.
GeForce RTX 5090 · 32 GB
Room to spare. 11.9 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. 5.1 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 12 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
- $0.070 in / $0.34 out
- Context served
- 262K
- Throughput
- Not measured
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| Cloudflare Workers AI | $0.10 / $0.30 | 256K | 69 tok/s | No | Yesunknown period | Unknown |
| OpenRouter | $0.070 / $0.34 | 262K | not measured | Unknown | Unknown | Unknown |
| DeepInfrafp8 | $0.070 / $0.34 | 262K | not measured | Unknown | Unknown | Unknown |
| DeepInfrafp8 | $0.070 / $0.34 | 262K | 23 tok/s | No | No | Confirmed |
| Venice AIbf16 | $0.13 / $0.40 | 256K | 20 tok/s | No | No | Confirmed |
| Parasailbf16 | $0.13 / $0.40 | 262K | 25 tok/s | No | No | Confirmed |
| NextBitbf16 | $0.12 / $0.40 | 262K | 52 tok/s | No | No | Confirmed |
| Ionstreambf16 | $0.13 / $0.40 | 262K | 19 tok/s | No | No | Confirmed |
| Novita AI | $0.13 / $0.40 | 262K | not measured | Unknown | Unknown | Unknown |
| Novita AIbf16 | $0.13 / $0.40 | 262K | 29 tok/s | No | No | Confirmed |
| SiliconFlowfp8 | $0.12 / $0.40 | 262K | 36 tok/s | No | No | Confirmed |
| Google Vertex AIglobal | $0.15 / $0.60 | 262K | 41 tok/s | No | No | Confirmed |
Cloudflare Workers AI
DeepInfra
Novita AI
Google Vertex AIAcross the 12 listings we hold: 9 say they do not train on prompts, 0 say they do and 3 do not say. 8 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 |
|---|---|---|---|
| Cloudflare Workers AI | ✓ | ✓ | ✗ |
| OpenRouter | ✓ | ✓ | ✓ |
| DeepInfrafp8 | |||
| DeepInfrafp8 | ✓ | ✓ | ✓ |
| Venice AIbf16 | ✓ | ✓ | ✓ |
| Parasailbf16 | ✗ | ✓ | ✓ |
| NextBitbf16 | ✓ | ✓ | ✓ |
| Ionstreambf16 | ✓ | ✓ | ✓ |
| Novita AI | |||
| Novita AIbf16 | ✓ | ✓ | ✗ |
| SiliconFlowfp8 | ✗ | ✓ | ✓ |
| Google Vertex AIglobal | ✓ | ✓ | ✓ |
Tool calling: 8 of 12 listings say yes, 2 say no, 2 publish no parameter list. JSON output: 10 of 12 listings say yes, 2 publish no parameter list. Strict schema: 8 of 12 listings say yes, 2 say no, 2 publish no parameter list.
Models people weigh against Gemma 4 26B A4B
When we formed this view
Dates behind this page
Prices last checked 5h 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 12 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 12 listings do not say whether they train on prompts.
Licence and identifiers
What the licence allowsApache License 2.0, 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
Apache License 2.0
Fully permissive: commercial use, redistribution, and derivatives allowed. Requires attribution and a copy of the license. Includes an express patent grant.
Identifiers
- Hugging Face
- google/gemma-4-26B-A4B-it
- Architecture
- Mixture of experts
- Modality record
- text+image+video->text
- Catalogue slug
- google-gemma-4-26b-a4b