Moonshot v1 vs Ministral 3 8B

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

Moonshot v1 is clearly ahead on the aggregate, 47 to 32. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.

Moonshot v1's sharpest advantage is in multimodal & grounded, where it averages 52.6 against 32.4. The single biggest benchmark swing on the page is MMLU, 53 to 28.

Quick Verdict

Pick Moonshot v1 if you want the stronger benchmark profile. Ministral 3 8B only becomes the better choice if its workflow or ecosystem matters more than the raw scoreboard.

Agentic

Moonshot v1

Moonshot v1

42.2

Ministral 3 8B

28.9

39
Terminal-Bench 2.0
26
49
BrowseComp
36
41
OSWorld-Verified
27

Coding

Moonshot v1

Moonshot v1

26.4

Ministral 3 8B

14.2

45
HumanEval
23
34
SWE-bench Verified
16
21
LiveCodeBench
15
30
SWE-bench Pro
13

Multimodal & Grounded

Moonshot v1

Moonshot v1

52.6

Ministral 3 8B

32.4

49
MMMU-Pro
27
57
OfficeQA Pro
39

Reasoning

Moonshot v1

Moonshot v1

55.5

Ministral 3 8B

36.1

51
SimpleQA
28
49
MuSR
26
73
BBH
63
58
LongBench v2
38
56
MRCRv2
41

Knowledge

Moonshot v1

Moonshot v1

42.3

Ministral 3 8B

28

53
MMLU
28
52
GPQA
27
50
SuperGPQA
25
48
OpenBookQA
23
64
MMLU-Pro
52
5
HLE
3
49
FrontierScience
32

Instruction Following

Moonshot v1

Moonshot v1

77

Ministral 3 8B

69

77
IFEval
69

Multilingual

Moonshot v1

Moonshot v1

69.8

Ministral 3 8B

61.7

73
MGSM
63
68
MMLU-ProX
61

Mathematics

Moonshot v1

Moonshot v1

61

Ministral 3 8B

43.3

53
AIME 2023
30
55
AIME 2024
32
54
AIME 2025
33
49
HMMT Feb 2023
26
51
HMMT Feb 2024
28
50
HMMT Feb 2025
27
52
BRUMO 2025
29
72
MATH-500
60

Frequently Asked Questions

Which is better, Moonshot v1 or Ministral 3 8B?

Moonshot v1 is ahead overall, 47 to 32. The biggest single separator in this matchup is MMLU, where the scores are 53 and 28.

Which is better for knowledge tasks, Moonshot v1 or Ministral 3 8B?

Moonshot v1 has the edge for knowledge tasks in this comparison, averaging 42.3 versus 28. Inside this category, MMLU is the benchmark that creates the most daylight between them.

Which is better for coding, Moonshot v1 or Ministral 3 8B?

Moonshot v1 has the edge for coding in this comparison, averaging 26.4 versus 14.2. Inside this category, HumanEval is the benchmark that creates the most daylight between them.

Which is better for math, Moonshot v1 or Ministral 3 8B?

Moonshot v1 has the edge for math in this comparison, averaging 61 versus 43.3. Inside this category, AIME 2023 is the benchmark that creates the most daylight between them.

Which is better for reasoning, Moonshot v1 or Ministral 3 8B?

Moonshot v1 has the edge for reasoning in this comparison, averaging 55.5 versus 36.1. Inside this category, SimpleQA is the benchmark that creates the most daylight between them.

Which is better for agentic tasks, Moonshot v1 or Ministral 3 8B?

Moonshot v1 has the edge for agentic tasks in this comparison, averaging 42.2 versus 28.9. Inside this category, OSWorld-Verified is the benchmark that creates the most daylight between them.

Which is better for multimodal and grounded tasks, Moonshot v1 or Ministral 3 8B?

Moonshot v1 has the edge for multimodal and grounded tasks in this comparison, averaging 52.6 versus 32.4. Inside this category, MMMU-Pro is the benchmark that creates the most daylight between them.

Which is better for instruction following, Moonshot v1 or Ministral 3 8B?

Moonshot v1 has the edge for instruction following in this comparison, averaging 77 versus 69. Inside this category, IFEval is the benchmark that creates the most daylight between them.

Which is better for multilingual tasks, Moonshot v1 or Ministral 3 8B?

Moonshot v1 has the edge for multilingual tasks in this comparison, averaging 69.8 versus 61.7. Inside this category, MGSM is the benchmark that creates the most daylight between them.

Last updated: March 12, 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.