Mistral Small 4
Mistral AI · released Jan 23, 2026 · mistralai/Mistral-Small-4-119B-2603
- Type
- Open weightsApache License 2.0
- Params
- 119B
- Context
- 262K
about 197K words of context
Our take
Written Sep 2, 2026Mistral Small 4 is a 119.4-billion-parameter text-and-image model released in January 2026 under a permissive Apache licence. It can handle up to 262,144 tokens in a single request, making it suited to long-document work, though no benchmark scores have been measured yet.
Pick this when you need a permissive licence for commercial or research deployment, or for processing long documents and image-based documents at scale. Skip it if you need validated quality scores, predictable throughput, or clarity on whether the full parameter count is active per token.
The case for it
- Apache 2.0 licence allows commercial use, modification and redistribution without restriction.
- 262,144-token request limit is among the largest we hold for this model, suited to document-scale work.
- Cheapest tracked rate sits well below the most expensive offer for both input and output.
The case against it
- No benchmark scores in our data — no measured chat, reasoning or coding quality at all.
- Active parameter count undisclosed; whether architecture is dense or mixture-of-experts and the true per-token compute cost remain unknown.
- Throughput varies by two-thirds even on the same provider's own API, from 52 to 86 tokens per second.
How good is it?
We hold no score for this model.
So there is no figure here for everyday use, coding, agent work or writing. That is a gap in our data, not a low score.
Can you run it yourself?
GeForce RTX 4090 · 24 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Apple M1 Pro (16-core GPU) · 32 GB
Too large for this card. The weights do not fit even with part of them offloaded to system memory.
Comfortable fit
Apple M1 Ultra (64-core GPU) · 128 GB
Room to spare. 16.1 GB spare means a 10% error in the size would not change the answer.
Memory use by level
Against a 24 GB card.
What is quantisation? →This model on every device we track71 devicesThe Q4 build most people download, on each device: what the weights come to, how much context the memory leaves, and whether it runs. Smallest device that runs it first.
Check against your own machine → · Where to rent it hosted →
Or rent it from someone else
Prices checked 2 hours ago — each listing carries its own date.
- per 1M tokens
- $0.15 in / $0.60 out
- Context served
- 262K
- Throughput
- ~98 tok/s
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| OpenRouterOpenRouter's own listing | $0.15 / $0.60checked 2 hours ago | 262K | not measured | Unknown | Unknown | Unknown |
| Mistral AIThrough OpenRouter | $0.15 / $0.60checked 2 hours ago | 262K210K max reply | 98 tok/s | No | Yes30 days | Confirmed |
| Mistral AIeuThrough OpenRouter | $0.17 / $0.66checked 2 hours ago | 262K210K max reply | 133 tok/s | No | Yes30 days | Confirmed |
| Mistral AIusThrough OpenRouter | $0.17 / $0.66checked 2 hours ago | 262K210K max reply | 86 tok/s | No | Yes30 days | Confirmed |
OpenRouterAcross the 4 listings we hold: 3 say they do not train on prompts, 0 say they do and 1 does not say. 3 appear in the zero-retention registry we check; the rest are unknown to us.
What each host's API supports
From the parameter list each endpoint publishes. Streaming is omitted: nothing we hold reports it, for any model.
| Provider | Tool calling | JSON output | Strict schema |
|---|---|---|---|
| OpenRouterOpenRouter's own listing | ✓ | ✓ | ✓ |
| Mistral AIThrough OpenRouter | ✓ | ✓ | ✓ |
| Mistral AIeuThrough OpenRouter | ✓ | ✓ | ✓ |
| Mistral AIusThrough OpenRouter | ✓ | ✓ | ✓ |
Tool calling: 4 of 4 listings say yes. JSON output: 4 of 4 listings say yes. Strict schema: 4 of 4 listings say yes.
Models people weigh against Mistral Small 4
When we formed this view
Recent changes
What moved
first indexed by our pipelineEach date is the day we first saw the change, or the day the maker announced it.
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.
- No independent board has scored it, so we hold no quality figures at all.
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- 1 of 4 listings does not say whether it trains on prompts.
- We hold no batch or off-peak rate for any of its listings.
- We hold a decode speed for it, but no prompt-processing (prefill) figure, so how long the input side of a job takes is unknown to us.
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
- mistralai/Mistral-Small-4-119B-2603
- Architecture
- Dense
- Takes in, gives back
- Text and images in, text out
- Catalogue slug
- mistralai-mistral-small-4