Mistral Small 4
Mistral AIfamily · Mistraldocs.mistral.ai/models/mistral-small-4-0-26-03
Hybrid model unifying instruct, reasoning, and coding in a single efficient model.
Updated 6 h ago · first seen 11 Sept 2026
model_01M2943ZSKTR767DHTCMVEKMB0
Overview
Identity
- Canonical model
- Yesidentity confidence: mediumOne row per real model release. Artifacts (checkpoints, quantisations, conversions) and folded evaluation variants point here.
- Official checkpoints
- official_checkpoints = hf_repo identifiers carried by the model itself; artifacts are separate entities pointing here through canonical_id.
- Artifacts
- None recordedSeparate entities (checkpoint · quantization · conversion · packaging) pointing to this model through canonical_id.
- Provider deployments
- 2
- API aliases
- mistral-small-4mistral-small-4-non-reasoningmistralai/mistral-small-2603Identifiers under which providers and evaluators refer to this model.
- Folded evaluation variants
- 1Effort / thinking variants (…-high, …-non-reasoning) are result configurations of this model, not separate models. Their old URLs redirect here.
Openness
Open weights— weights downloadable under Apache-2.0; commercial use allowed; redistribution allowed; derivatives allowed; 4 dimensions unknown.
Weights downloadable under a permissive or Creative Commons licence allowing commercial use; code or data may be missing.
Weights
Yes
Inference code
—
Training code
—
Training data
—
Dataset
—
Commercial use
Yes
Redistribution
Yes
Derivatives
Yes
Licence: Apache License 2.0 (permissive · SPDX Apache-2.0 · stated as “Apache 2.0”)
dimensions marked null are unknown, not false
Key facts
- Release date
Source:OpenRouter public model & pricing listingT2observed 5 d agomedium
- Version
Source:Mistral AI docsT1observed 5 d agohigh
- Official page
Source:Mistral AI docsT1observed 5 d agohigh
- Openrouter id
Source:OpenRouter public model & pricing listingT2observed 5 d agomedium
Architecture
- Tokenizer
Source:OpenRouter public model & pricing listingT2observed 5 d agomedium
- Hugging Face repo
Source:OpenRouter public model & pricing listingT2observed 5 d agomedium
Capabilities
Modalities
- Modalities
- imagetext
- Input
- imagetext
- Output
- text
Capabilities
Tool calling
Yes
Mistral AI docs · T1
Structured output
Yes
Mistral AI docs · T1
Reasoning
Yes
Artificial Analysis · T2
Vision
Yes
OpenRouter public model & pricing listing · T2
Audio
Unavailable
Fine-tuning available
Unavailable
- Context window
Source:Mistral AI docsT1observed 36 h agohigh
- Max output
Source:OpenRouter public model & pricing listingT2observed 5 d agomedium
- Tokenizer
Source:OpenRouter public model & pricing listingT2observed 5 d agomedium
Benchmarks34
Compare with another model →Comparable same task and conditions · Partially comparable same task, conditions differ (effort, temperature, judge) · Not comparable different variant or metric
Current rows only, grouped by benchmark → canonical metric → comparability group (task configuration). Effort variants folded into this model appear as rows of the same group. 34 current rows in total. “vs leader” compares with the current leader of the benchmark's primary group only; other groups are not directly comparable. Comparability rules →
Providers & Pricing5
All offers in the price terminal →USD per 1M tokens as published by each provider; native units (per-request fees, flex/priority tiers) are kept verbatim. Rows are append-only — every price change is kept in the history below. Cost of a workload →
Price history
Output price · USD / 1M tokens 2 providers
- Mistral AI La Plateforme
- OpenRouter
- OpenRouter$0.60 → $0.3012 Sept 2026
- OpenRouterfirst observed $0.6012 Sept 2026
- Mistral AI La Plateforme$0.60 → $0.3011 Sept 2026
- Mistral AI La Plateformefirst observed $0.6011 Sept 2026
Input price · USD / 1M tokens 2 providers
- Mistral AI La Plateforme
- OpenRouter
- OpenRouter$0.15 → $0.07512 Sept 2026
- OpenRouterfirst observed $0.1512 Sept 2026
- Mistral AI La Plateforme$0.15 → $0.07511 Sept 2026
- Mistral AI La Plateformefirst observed $0.1511 Sept 2026
Versions & Artifacts0
Version history
Context window2 changes
11 Sept 2026→12 Sept 2026→15 Sept 2026current
Licensefirst observation only
11 Sept 2026current
Max outputfirst observation only
11 Sept 2026current
Opennessfirst observation only
11 Sept 2026current
Each hop is a claim: click a value for its source, tier and observation time. Nothing is overwritten — a new observation closes the previous claim.
Artifacts 0
No artifact (checkpoint, quantisation, conversion or packaging) points to this model yet.
Timeline23
Full timeline →Mistral Small 4: context length changed from 262144 to 256000
Context window262.1K tokens→256K tokensmistralOpenRouter lists Mistral Small 4 at $0.075 in / $0.3 out per 1M tokens
openrouterOpenRouter lists Mistral Small 4 at $0.15 in / $0.6 out per 1M tokens
openrouterMistral Small 4: context length changed from 256000 to 262144
Context window256K tokens→262.1K tokensopenrouterMistral Small 4 scores 10.61% on Terminal-Bench
artificial_analysisMistral Small 4 scores 18.42% on τ²-bench
artificial_analysisMistral Small 4 scores 46.47% on MMMU-Pro
artificial_analysisMistral Small 4 scores 32.79% on IFBench
artificial_analysisMistral Small 4 scores 3.8% on Humanity's Last Exam
artificial_analysisMistral Small 4 scores 57.07% on GPQA Diamond
artificial_analysisMistral Small 4 scores 8.99 on Artificial Analysis Intelligence Index
artificial_analysisMistral Small 4 scores 17.42% on Terminal-Bench
artificial_analysisMistral Small 4 scores 20.97% on Terminal-Bench
artificial_analysisMistral Small 4 scores 0% on Terminal-Bench
artificial_analysisMistral Small 4 scores 41.23% on τ²-bench
artificial_analysisMistral Small 4 scores 56.82% on MMMU-Pro
artificial_analysisMistral Small 4 scores 48.16% on IFBench
artificial_analysisMistral Small 4 scores 38.77% on SciCode
artificial_analysisMistral Small 4 scores 9.87% on Humanity's Last Exam
artificial_analysisMistral Small 4 scores 76.87% on GPQA Diamond
artificial_analysisMistral Small 4 scores 11.45 on Artificial Analysis Intelligence Index
artificial_analysisMistral Small 4: reasoning changed from false to true
ReasoningNo→Yesartificial_analysisMistral AI La Plateforme lists Mistral Small 4 at $0.15 in / $0.6 out per 1M tokens
mistral
Change history54
Viewing AI Atlas as of 15 Sept 2026 — attributes exactly as the atlas knew them on that day; later corrections are not shown.
Back to today →Attributes as of 15 Sept 2026 32 claims in force
- Version
- 26.03
- Release date
- 16 Mar 2026
- Openness
- open-weights
- License
- Apache-2.0
- Context window
- 256K tokens
- Max output
- 209.7K tokens
- Modalities
- image, text
- Input modalities
- image, text
- Output modalities
- text
- Tokenizer
- Mistral
- API alias
- mistral-small-latest
- Hugging Face repo
- mistralai/Mistral-Small-4-119B-2603
- Official page
- Aa context window
- 256,000
- Aa openness
- open-weights
- Api aliases
- mistral-small-latest
- Capabilities
- function_calling, structured_output
- Commercial use allowed
- Yes
- Derivatives allowed
- Yes
- License key
- Apache-2.0
- License raw
- Apache 2.0
- Aa median output tokens per second
- 159.7
- Openrouter id
- mistralai/mistral-small-2603
- Openrouter listed at
- 16 Mar 2026
- Reasoning
- Yes
- Redistribution allowed
- Yes
- Structured output
- Yes
- Supported parameters
- frequency_penalty, include_reasoning, max_tokens, presence_penalty, reasoning, reasoning_effort, response_format, seed, stop, structured_outputs, temperature, tool_choice, tools, top_k, top_p
- Tool calling
- Yes
- Vision
- Yes
- Weights available
- Yes
Versionversion1
Release daterelease_date1
Opennessopenness2
Licenselicense1
Context windowcontext_length11conflicting claims
Max outputmax_output_tokens1
Modalitiesmodalities1
Input modalitiesmodalities_input1
Output modalitiesmodalities_output1
Tokenizertokenizer1
API aliasapi_alias1
Hugging Face repohf_repo1
Official pageofficial_url1
Aa context windowaa_context_window1
Aa opennessaa_openness1
Api aliasesapi_aliases1
Capabilitiescapabilities1
Commercial use allowedcommercial_use_allowed1
Derivatives allowedderivatives_allowed1
Descriptiondescription1
License keylicense_key1
License rawlicense_raw1
Aa median output tokens per secondmetric.aa_median_output_tokens_per_second10
Openrouter idopenrouter_id1
Openrouter listed atopenrouter_listed_at1
Reasoningreasoning3
Redistribution allowedredistribution_allowed1
Structured outputstructured_output1
Supported parameterssupported_parameters1
Tool callingtool_calling1
Visionvision1
Weights availableweights_available1
Claims are temporal and append-only: a new observation closes the previous claim (valid_to) instead of overwriting it. Conflicting claims from different sources are kept side by side and flagged — never averaged. Methodology →
Provenance
Attributed facts
33
Source tiers
T1T213 / 20
Freshest observation
6 h ago
Conflicts
7 flagged
Source documents 4
Tier 1 = official/primary, 2 = quality secondary, 3 = community, 4 = unverified. Every snapshot is archived; see all sources and the methodology.
Data quality (76/100) measures how well AI Atlas knows this entity — completeness, primary-source ratio, freshness, conflicts — never how good the model is. Methodology →