Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
LFM2-24B-A2B is clearly ahead on the aggregate, 38 to 33. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
LFM2-24B-A2B's sharpest advantage is in coding, where it averages 18 against 8.2. The single biggest benchmark swing on the page is HumanEval, 42 to 17. LFM2.5-1.2B-Thinking does hit back in instruction following, so the answer changes if that is the part of the workload you care about most.
LFM2-24B-A2B is also the more expensive model on tokens at $0.03 input / $0.12 output per 1M tokens, versus $0.00 input / $0.00 output per 1M tokens for LFM2.5-1.2B-Thinking. That is roughly Infinityx on output cost alone. LFM2.5-1.2B-Thinking is the reasoning model in the pair, while LFM2-24B-A2B is not. That usually helps on harder chain-of-thought-heavy tests, but it can also mean more latency and more token spend in real use.
Pick LFM2-24B-A2B if you want the stronger benchmark profile. LFM2.5-1.2B-Thinking only becomes the better choice if instruction following is the priority or you want the cheaper token bill.
LFM2-24B-A2B
33.4
LFM2.5-1.2B-Thinking
34.1
LFM2-24B-A2B
18
LFM2.5-1.2B-Thinking
8.2
LFM2-24B-A2B
41.7
LFM2.5-1.2B-Thinking
32.4
LFM2-24B-A2B
46.6
LFM2.5-1.2B-Thinking
38.4
LFM2-24B-A2B
35.6
LFM2.5-1.2B-Thinking
27
LFM2-24B-A2B
68
LFM2.5-1.2B-Thinking
72
LFM2-24B-A2B
61.4
LFM2.5-1.2B-Thinking
60.7
LFM2-24B-A2B
50.4
LFM2.5-1.2B-Thinking
42.3
LFM2-24B-A2B is ahead overall, 38 to 33. The biggest single separator in this matchup is HumanEval, where the scores are 42 and 17.
LFM2-24B-A2B has the edge for knowledge tasks in this comparison, averaging 35.6 versus 27. Inside this category, MMLU is the benchmark that creates the most daylight between them.
LFM2-24B-A2B has the edge for coding in this comparison, averaging 18 versus 8.2. Inside this category, HumanEval is the benchmark that creates the most daylight between them.
LFM2-24B-A2B has the edge for math in this comparison, averaging 50.4 versus 42.3. Inside this category, AIME 2023 is the benchmark that creates the most daylight between them.
LFM2-24B-A2B has the edge for reasoning in this comparison, averaging 46.6 versus 38.4. Inside this category, SimpleQA is the benchmark that creates the most daylight between them.
LFM2.5-1.2B-Thinking has the edge for agentic tasks in this comparison, averaging 34.1 versus 33.4. Inside this category, Terminal-Bench 2.0 is the benchmark that creates the most daylight between them.
LFM2-24B-A2B has the edge for multimodal and grounded tasks in this comparison, averaging 41.7 versus 32.4. Inside this category, MMMU-Pro is the benchmark that creates the most daylight between them.
LFM2.5-1.2B-Thinking has the edge for instruction following in this comparison, averaging 72 versus 68. Inside this category, IFEval is the benchmark that creates the most daylight between them.
LFM2-24B-A2B has the edge for multilingual tasks in this comparison, averaging 61.4 versus 60.7. Inside this category, MGSM 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.