Mistral Large 3 vs Mixtral 8x22B Instruct v0.1

Side-by-side benchmark comparison across knowledge, coding, math, and reasoning.

Mistral Large 3 is clearly ahead on the aggregate, 68 to 28. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.

Mistral Large 3 is also the more expensive model on tokens at $2.00 input / $6.00 output per 1M tokens, versus $0.00 input / $0.00 output per 1M tokens for Mixtral 8x22B Instruct v0.1. That is roughly Infinityx on output cost alone. Mistral Large 3 gives you the larger context window at 128K, compared with 64K for Mixtral 8x22B Instruct v0.1.

Quick Verdict

Pick Mistral Large 3 if you want the stronger benchmark profile. Mixtral 8x22B Instruct v0.1 only becomes the better choice if knowledge is the priority or you want the cheaper token bill.

Knowledge

Mixtral 8x22B Instruct v0.1

Mistral Large 3

63.5

Mixtral 8x22B Instruct v0.1

71.4

76
MMLU
71.4
75
GPQA
-
73
SuperGPQA
-
71
OpenBookQA
-
74
MMLU-Pro
-
12
HLE
-

Coding

Mixtral 8x22B Instruct v0.1

Mistral Large 3

50.7

Mixtral 8x22B Instruct v0.1

54.8

68
HumanEval
54.8
45
SWE-bench Verified
-
39
LiveCodeBench
-

Mathematics

Mistral Large 3
76
AIME 2023
-
78
AIME 2024
-
77
AIME 2025
-
72
HMMT Feb 2023
-
74
HMMT Feb 2024
-
73
HMMT Feb 2025
-
75
BRUMO 2025
-
80
MATH-500
-

Reasoning

Mistral Large 3
73
SimpleQA
-
71
MuSR
-
81
BBH
-

Instruction Following

Mistral Large 3
83
IFEval
-

Multilingual

Mistral Large 3
82
MGSM
-

Frequently Asked Questions

Which is better, Mistral Large 3 or Mixtral 8x22B Instruct v0.1?

Mistral Large 3 is ahead overall, 68 to 28. The biggest single separator in this matchup is HumanEval, where the scores are 68 and 54.8.

Which is better for knowledge tasks, Mistral Large 3 or Mixtral 8x22B Instruct v0.1?

Mixtral 8x22B Instruct v0.1 has the edge for knowledge tasks in this comparison, averaging 71.4 versus 63.5. Inside this category, MMLU is the benchmark that creates the most daylight between them.

Which is better for coding, Mistral Large 3 or Mixtral 8x22B Instruct v0.1?

Mixtral 8x22B Instruct v0.1 has the edge for coding in this comparison, averaging 54.8 versus 50.7. Inside this category, HumanEval is the benchmark that creates the most daylight between them.

Last updated: March 9, 2026

Weekly LLM Benchmark Digest

Get notified when new models drop, benchmark scores change, or the leaderboard shifts. One email per week.

Free. No spam. Unsubscribe anytime. We only store derived location metadata for consent routing.