LLMcompare

Search

Search for a command to run...

A · open · Mistral

Mistral Large 3

Mistral

B · open · Mistral

Mistral Medium 3.5

Mistral

2:10

overall capability wins (2 ties across 14 scored). Mistral Medium 3.5 leads overall.

  • Reasoning1:5
  • Coding0:3
  • Arena0:1
  • Tool use0:1
  • Composite indices0:2
  • Pricing1:1
  • Speed0:2
  • Specs1:0

Verdict

Mistral Medium 3.5 leads coding (widest gap: +38.6 on TermBench); Mistral Medium 3.5 ranks higher on LMArena (+13 Elo).

  • Mistral Medium 3.5 leads coding (widest gap: +38.6 on TermBench)
  • Mistral Medium 3.5 ranks higher on LMArena (+13 Elo)

Cost for 1M in + 250K out

Illustrative chat workload at primary-provider list prices.

Mistral Large 3
$3.50
Mistral Medium 3.5
$3.38

Mistral Medium 3.5 is about 1.0x cheaper on this workload.

Price per 1M tokens

Full comparison

Organization

Mistral

Mistral

Family

Mistral

Mistral

License

Apache 2.0

Modified MIT

Open weights

Yes

Yes

Release

Nov 19, 2025

Apr 22, 2026

Knowledge cutoff

-

-

API / provider

Mistral

Mistral

Modalities

text, image → text

text, image → text

Specs

Context window

256K

256K

tie

Max output

64K

64K

tie

Parameters

675B (41B act.)

128B

A +547

Pricing

Input $/1M

$2

$1.50

B +$0.50

Output $/1M

$6

$7.50

A +$1.50

Blended $/1M (3∶1)

$3

$3

tie

Speed

tok/s

70

80

B +10

TTFT (s)

0.45

0.4

B +0.05

Reasoning

MMLU-Pro

80.7

—

—

GPQA Diamond

68.0

74.8

B +6.8 pts

Humanity's Last Exam

4.2

13.8

B +9.6 pts

AIME 2025

38.0

—

—

MATH-500

—

—

—

Humanity's Last Exam (with tools)

—

—

—

AA-Omniscience Accuracy

25.0

24.7

A +0.3 pts

AA-LCR v1.1

36.0

69.3

B +33.3 pts

CritPt

—

—

—

MMMU-Pro

55.7

64.9

B +9.2 pts

IFBench

36.2

68.8

B +32.6 pts

Chartography

—

—

—

Chartography (With Tools)

—

—

—

Coding

SWE-bench Verified

—

—

—

SWE-bench Pro

—

—

—

SWE-bench Multilingual

—

—

—

LiveCodeBench

46.5

—

—

Terminal-Bench 2.1

12.0

50.6

B +38.6 pts

Aider Polyglot

—

—

—

Terminal-Bench 3

—

—

—

BigCodeBench

—

—

—

SciCode

36.6

40.2

B +3.6 pts

CursorBench

—

—

—

SWE-Rebench

—

—

—

NL2Repo-Bench

—

—

—

DeepSWE

—

—

—

WebDev Arena

1,230

1,265

B +35 Elo

Terminal-Bench 4.0

—

—

—

CursorBench 4.0

—

—

—

FrontierCode v1.1 (Main)

—

—

—

Arena

LMArena Elo

1,413

1,426

B +13 Elo

Mistral Large 3

Context
256K / 64K out
Parameters
675B (41B act.)
Price
$2 / $6
Speed
70 tok/s · 0.45s TTFT
Modalities
text, image → text
License
Apache 2.0

Mistral's open Large 3 MoE - multilingual European frontier alternative under Apache 2.0.

Mistral Medium 3.5

Context
256K / 64K out
Parameters
128B
Price
$1.50 / $7.50
Speed
80 tok/s · 0.4s TTFT
Modalities
text, image → text
License
Modified MIT

Mistral's April 2026 frontier multimodal model (open weights, Modified MIT) powering Vibe remote agents at $1.50/$7.50.

Benchmark charts

Winner bars are emphasized. Per-benchmark deltas sit above each chart.

Reasoning & knowledge

GPQA: B +6.8HLE: B +9.6AA-LCR: B +33.3Omniscience Acc.: A +0.3MMMU-Pro: B +9.2IFBench: B +32.6

Coding

TermBench: B +38.6SciCode: B +3.6WebDev Elo: B +35

Tool use & function calling

τ³-Banking: B +9.3

Arena

Arena Elo: B +13

Composite indices

AA Index: B +3.8AA Index v4.3: B +5.2

Pick a different pair · Back to catalog