Models/Pixtral Large

MDL

Pixtral Large

Mistral

Verified against Artificial Analysis · Jul 13, 2026

Capability

percentile index

Price /M

$2 / $6

in / out

Context

—

Avg score

33.5

10 benchmarks

AA Intelligence

8.1

index

Weights

Closed

Released

Nov 18, 2024

About

Pixtral Large is a proprietary model in the Pixtral family from Mistral. Benchmarked on 10 evals, averaging 33.5.

Benchmark scores · 10

avg 33.5 — higher is better

MATHartificial-analysis · 2026-07

71.4

MMLU-Proartificial-analysis · 2026-07

70.1

GPQA Diamondartificial-analysis · 2026-07

50.5

τ²-benchartificial-analysis · 2026-07

36.5

IFBenchartificial-analysis · 2026-07

34.5

SciCodeartificial-analysis · 2026-07

29.2

LiveCodeBenchartificial-analysis · 2026-07

26.1

AA Long-Context Reasoningartificial-analysis · 2026-07

10.3

Humanity's Last Examartificial-analysis · 2026-07

3.6

AIMEartificial-analysis · 2026-07

2.3

The research behind this model

all papers mentioning it →

Same Model, Different Weakness: How Language and Modality Reshape the Jailbreak Attack Surface in Frontier MLLMsMay 22, 2026

Mentions matched by name in title/abstract — from the arXiv + HF daily corpus.

Details

PublisherMistral

FamilyPixtral

Parameters—

License—

WeightsClosed

ReleasedNov 18, 2024

Performance & price

Price in / out$2 / $6 per 1M

Blended price$3 / 1M

Math index2.3

Median across API providers, via Artificial Analysis. Blended = 3:1 input:output. For reasoning models, latency includes thinking time. Methodology →

History

full ledger →

Capability over time

Append-only ledger — every observed change to this model's numbers.

API providers

Artificial Analysis$6/M