Models/Mistral/Mistral Small 4

Mistral Small 4

Mistral Small 4 is the next major release in the Mistral Small family, unifying the capabilities of several flagship Mistral models into a single system. It combines strong reasoning from...

01

Identity

262,144Context in
Not published on OpenRouterContext out
2026-03-16 (OpenRouter created 1773695685)Released
OpenWeights
Canonical name
Mistral Small 4
Aliases
mistralai/mistral-small-2603, mistralai/Mistral-Small-4-119B-2603
Continuum hosted id
Not listed as a Continuum hosted id
OpenRouter slug
mistralai/mistral-small-2603
Hugging Face
mistralai/Mistral-Small-4-119B-2603
Modalities
text, image
Open weights Not on Continuum host list Open

Context window on the 2026-08-19 OpenRouter row: 262,144 tokens.

Max completion tokens were not listed on that row.

Input / output on that row: $0.15 / $0.60 per 1M tokens.

Architecture fields: text+image->text · Mistral.

Hugging Face id on that row: mistralai/Mistral-Small-4-119B-2603.

02

Should I use this for coding agents

Mistral Small 4 is the next major release in the Mistral Small family, unifying the capabilities of several flagship Mistral models into a single system.

When you want the 2026-03-16 catalog SKU, not a later rename.

No independent row for this SKU

No independent board lists this exact SKU as of 2026-08-19.

No other card from this lab carries an independent board row either, so there is no sibling worth pointing at.

Boards checked: DeepSWE official (2026-08-13), Artificial Analysis Intelligence Index, vals.ai hero index. Every card that does carry a row is on the coding leaderboard.

DeepSWE

No official DeepSWE mini-swe-agent row for this identity on the public board we fetched.

omitted, not invented as of 2026-08-13 source
Artificial Analysis

No Intelligence Index integer fetched for this identity. Chip omitted.

omitted, not invented as of 2026-08-19 source
Vals Index

We did not open a vals.ai card for this identity. Chip omitted.

omitted, not invented as of 2026-08-19 source
03

When not to use it

Trust this before you buy

  • When you need a Continuum-hosted SKU from this lab, use the hosted card on this hub instead of this catalog row.
  • $0.15 in / $0.60 out per 1M on the 2026-08-19 OpenRouter row. Context 262,144 in.
  • Need a denser published sibling on this hub? Mistral Large 2512 is the card already on the cluster.
  • When the job needs a published DeepSWE, Artificial Analysis, or vals.ai chip: this page does not invent one. Those boards did not list this exact SKU on 2026-08-19.

Plus is $25/mo with $25 weekly hosted usage. The Mac app stays free with your own keys. Get Plus is not Download for Mac.

04

Artifact and Hugging Face downloads

Artifact

SPDX / license
apache-2.0
Total params
HF safetensors.total: 119,401,317,952 stored tensors
Native precision
HF tensors BF16 + F8_E4M3
Repo size
HF API usedStorage 241,885,986,991 bytes
HF created
2026-01-23T13:14:14.000Z
Official HF repo
mistralai/Mistral-Small-4-119B-2603
HF downloads
93,051
HF likes
419

Official weights id on the 2026-08-19 OpenRouter row: mistralai/Mistral-Small-4-119B-2603. Community GGUF is a quant, not a second Continuum card.

Official weights

This is the model repo, not Download for Mac. Downloads and likes from the Hugging Face API on 2026-08-19.

Hub file tree

Safetensor shards
10
Shard bytes
241,850,185,032 bytes
Other file bytes
39,325,713 bytes
Tree file count
28
Hub usedStorage
241,885,986,991 bytes

usedStorage and the sum of .safetensors sizes can differ (LFS pointers, non-shard files, duplicate copies). Cited separately. Source: recursive tree API, 2026-08-19.

PathBytes
consolidated-00001-of-00007.safetensors20,000,632,534
consolidated-00002-of-00007.safetensors19,997,737,172
consolidated-00003-of-00007.safetensors19,997,736,948
consolidated-00004-of-00007.safetensors19,997,737,108
consolidated-00005-of-00007.safetensors19,997,738,276
consolidated-00006-of-00007.safetensors19,861,887,138
consolidated-00007-of-00007.safetensors1,073,741,920
model-00001-of-00003.safetensors49,078,433,296
model-00002-of-00003.safetensors49,132,710,896
model-00003-of-00003.safetensors22,711,829,744
tokenizer.json17,077,420
tekken.json16,275,354
consolidated.safetensors.index.json3,619,748
images/aime.png601,172
images/lcr.png586,377
images/livecode.png571,479
images/image3.png220,228
model.safetensors.index.json159,870
images/image2.png152,184
tokenizer_config.json21,140
README.md16,483
chat_template.jinja15,168
config.json2,436
SYSTEM_PROMPT.txt2,372
.gitattributes1,939
params.json1,510
processor_config.json701
generation_config.json132

All 28 files the tree API returned are listed: every one of the 10 safetensor shards plus 18 other files. Files tab: https://huggingface.co/mistralai/Mistral-Small-4-119B-2603/tree/main.

05

Continuum serving

OpenRouter slug mistralai/mistral-small-2603 on the 2026-08-19 catalog.

First party: https://docs.mistral.ai/getting-started/models/.

Not on the Continuum host list fetched 2026-08-19.

Continuum hosted id not on the host list

We list first-party and OpenRouter, plus Continuum only when the live public allowlist named the id. This is not a 15-host routing table.

06

Economics

OpenRouter 2026-08-19: $0.15 in / $0.60 out per 1M tokens. Context 262,144 in.

No separate internal-reasoning price on that row.

Price this model Opens the pricing calculator preloaded with Mistral Small 4.
07

Provenance

ClaimSourceAs of
OpenRouter id mistralai/mistral-small-2603; context 262144; created 1773695685; $0.15 / $0.60 per 1MOpenRouter /api/v1/models2026-08-19
OpenRouter id mistralai/mistral-small-2603, context 262,144OpenRouter /api/v1/models2026-08-19
HF downloads 93,051Hugging Face API mistralai/Mistral-Small-4-119B-26032026-08-19
HF safetensors.total: 119,401,317,952 stored tensors; HF API usedStorage 241,885,986,991 bytes; created 2026-01-23T13:14:14.000Z; HF tensors BF16 + F8_E4M3Hugging Face API mistralai/Mistral-Small-4-119B-26032026-08-19
10 safetensor shards; shard bytes 241,850,185,032; tree files 28Hugging Face tree API mistralai/Mistral-Small-4-119B-26032026-08-19
08

Compare, FAQ, and Get Plus

No other card in this cluster shares a published DeepSWE official row we can put next to this one. We will not compare on vendor-blog numbers.

Continue through Mistral's model family with Ministral 3 14B 2512.

FAQ

Why does the lab blog disagree with DeepSWE or Scale?

Lab posts pick a harness, an effort, a split, and sometimes a private eval set. DeepSWE publishes the official mini-swe-agent row with cost, tokens, and steps. Scale SWE-bench Pro only counts when the public shared-harness board has a row we can fetch. A higher lab number is usually a different test, not a better one.

OpenAI-compatible call

No Continuum hosted id is verified for this model, so this page does not invent a curl target. Use the first-party API or the OpenRouter slug mistralai/mistral-small-2603.

Plus is $25/mo with $25 weekly hosted usage. The Mac app stays free with your own keys. Get Plus is not Download for Mac.