Mistral Small 3
Mistral AI · released Jan 28, 2025 · mistralai/Mistral-Small-24B-Instruct-2501
- Type
- Open weightsApache License 2.0
- Params
- 23.6B
- Context
- 33K
about 25K words of context
Our take
Written Sep 5, 2026Mistral Small 3 is a 23.6-billion-parameter text model released in early 2025 with a permissive Apache licence. Its coding score is the strongest of its six measured skills, and hosted inference is available at low cost from a small set of providers.
Pick this for budget text generation where a permissive licence matters, or for coding assistance specifically — its coding score sits 38 points above its general text rating. Use it for self-hosted or redistributed deployment needing open weights and commercial terms. Skip it if you need multimodal input, creative writing quality, or evidence of mixture-of-experts efficiency.
The case for it
- Coding is its standout skill: 1312.3 on Arena Coding, 38.2 points above its overall text score.
- Apache 2.0 licence with low hosted cost on two tracked providers.
- DeepInfra discloses throughput at 35 tokens per second.
The case against it
- Creative writing is its weakest measured skill at 1226.4, 47.7 points below its overall text score and 85.9 below coding.
- No active-parameter count disclosed, so there is no evidence of efficiency gains from sparse architecture.
- Only three tracked offers, a thin hosting market.
How good is it?
An open-weights text model for general chat, though it trails most models on everyday questions, drafting and code.
- getting answers to everyday questionsArena Text (overall) · 155th of 168
- drafts, rewrites and editingArena Creative Writing · 154th of 168
- writing and completing codeArena Coding · 153rd of 168
EverydayGeneral questions and everyday reasoning
Arena Text (overall)155th of 168 · 1274
CodingWriting and fixing code on its own
Arena Coding153rd of 168 · 1312
Arena Coding is the only board that has scored it for this.
AgenticPlanning, calling tools, staying on task
Not yet scored on Arena Agent.
WritingDrafting and rewriting prose
Arena Creative Writing154th of 168 · 1227
Arena Creative Writing is the only board that has scored it for this.
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.
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?
Comfortable fit
GeForce RTX 4090 · 24 GB
Room to spare. 5.9 GB spare means a 10% error in the size would not change the answer.
GeForce RTX 5090 · 32 GB
Room to spare. 13.9 GB spare means a 10% error in the size would not change the answer.
Apple M2 (10-core GPU) · 24 GB
Borderline fit on an estimated size. It leaves 1.1 GB spare on a size we calculated rather than measured, and a 10% error either way would 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
- $0.050 in / $0.080 out
- Context served
- 33K
- Throughput
- Not measured
| Provider | In / out per 1M tokens | Context | Throughput | Trains on prompts | Logs prompts | Zero retention |
|---|---|---|---|---|---|---|
| OpenRouterOpenRouter's own listing | $0.050 / $0.080checked 2 hours ago | 33K | not measured | Unknown | Unknown | Unknown |
| DeepInfrafp8Direct and through OpenRouter | $0.050 / $0.080checked 2 hours ago | 33K16K max reply through OpenRouter | 42 tok/sthrough OpenRouter | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterNo | DirectUnknownThrough OpenRouterConfirmed |
DeepInfraAcross the 2 listings we hold: 1 says it does not train on prompts (only through OpenRouter), 0 say they do and 1 does not say. 1 appears in the zero-retention registry we check (only through OpenRouter); 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 | ✗ | ✓ | ✓ |
| DeepInfrafp8Direct and through OpenRouter | ✗ | ✓ | ✓ |
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 Mistral Small 3
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.
- Nothing we hold says whether an endpoint streams, so we do not show it either way.
- 1 of 2 listings does not say whether it trains on prompts, and 1 answers only through OpenRouter, not for its own listing.
- We hold no cached-input rate for any of its listings.
- 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-24B-Instruct-2501
- Architecture
- Dense
- Takes in, gives back
- Text in, text out
- Catalogue slug
- mistralai-mistral-small-3