Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
Claude Haiku 4.5 is clearly ahead on the aggregate, 62 to 55. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
Claude Haiku 4.5's sharpest advantage is in coding, where it averages 41.7 against 33. The single biggest benchmark swing on the page is OSWorld-Verified, 57 to 44.
Claude Haiku 4.5 is also the more expensive model on tokens at $0.80 input / $4.00 output per 1M tokens, versus $0.00 input / $0.00 output per 1M tokens for Ministral 3 14B. That is roughly Infinityx on output cost alone. Claude Haiku 4.5 gives you the larger context window at 200K, compared with 128K for Ministral 3 14B.
Pick Claude Haiku 4.5 if you want the stronger benchmark profile. Ministral 3 14B only becomes the better choice if you want the cheaper token bill.
Claude Haiku 4.5
56.7
Ministral 3 14B
48.4
Claude Haiku 4.5
41.7
Ministral 3 14B
33
Claude Haiku 4.5
78.4
Ministral 3 14B
70.5
Claude Haiku 4.5
68.9
Ministral 3 14B
63.6
Claude Haiku 4.5
53.6
Ministral 3 14B
50.1
Claude Haiku 4.5
86
Ministral 3 14B
80
Claude Haiku 4.5
80.1
Ministral 3 14B
76.8
Claude Haiku 4.5
73.3
Ministral 3 14B
69.7
Claude Haiku 4.5 is ahead overall, 62 to 55. The biggest single separator in this matchup is OSWorld-Verified, where the scores are 57 and 44.
Claude Haiku 4.5 has the edge for knowledge tasks in this comparison, averaging 53.6 versus 50.1. Inside this category, MMLU-Pro is the benchmark that creates the most daylight between them.
Claude Haiku 4.5 has the edge for coding in this comparison, averaging 41.7 versus 33. Inside this category, SWE-bench Pro is the benchmark that creates the most daylight between them.
Claude Haiku 4.5 has the edge for math in this comparison, averaging 73.3 versus 69.7. Inside this category, MATH-500 is the benchmark that creates the most daylight between them.
Claude Haiku 4.5 has the edge for reasoning in this comparison, averaging 68.9 versus 63.6. Inside this category, LongBench v2 is the benchmark that creates the most daylight between them.
Claude Haiku 4.5 has the edge for agentic tasks in this comparison, averaging 56.7 versus 48.4. Inside this category, OSWorld-Verified is the benchmark that creates the most daylight between them.
Claude Haiku 4.5 has the edge for multimodal and grounded tasks in this comparison, averaging 78.4 versus 70.5. Inside this category, MMMU-Pro is the benchmark that creates the most daylight between them.
Claude Haiku 4.5 has the edge for instruction following in this comparison, averaging 86 versus 80. Inside this category, IFEval is the benchmark that creates the most daylight between them.
Claude Haiku 4.5 has the edge for multilingual tasks in this comparison, averaging 80.1 versus 76.8. Inside this category, MMLU-ProX is the benchmark that creates the most daylight between them.
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.