Side-by-side benchmark comparison across agentic, coding, multimodal, knowledge, reasoning, and math workflows.
DeepSeek-R1 is clearly ahead on the aggregate, 43 to 30. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
DeepSeek-R1's sharpest advantage is in agentic, where it averages 44.5 against 25.7. The single biggest benchmark swing on the page is LongBench v2, 58 to 34. LFM2.5-1.2B-Instruct does hit back in instruction following, so the answer changes if that is the part of the workload you care about most.
DeepSeek-R1 is also the more expensive model on tokens at $0.55 input / $2.19 output per 1M tokens, versus $0.00 input / $0.00 output per 1M tokens for LFM2.5-1.2B-Instruct. That is roughly Infinityx on output cost alone. DeepSeek-R1 is the reasoning model in the pair, while LFM2.5-1.2B-Instruct 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. DeepSeek-R1 gives you the larger context window at 128K, compared with 32K for LFM2.5-1.2B-Instruct.
Pick DeepSeek-R1 if you want the stronger benchmark profile. LFM2.5-1.2B-Instruct only becomes the better choice if instruction following is the priority or you want the cheaper token bill.
DeepSeek-R1
44.5
LFM2.5-1.2B-Instruct
25.7
DeepSeek-R1
21.5
LFM2.5-1.2B-Instruct
7.2
DeepSeek-R1
47.5
LFM2.5-1.2B-Instruct
32.4
DeepSeek-R1
50.9
LFM2.5-1.2B-Instruct
32.1
DeepSeek-R1
37.9
LFM2.5-1.2B-Instruct
26
DeepSeek-R1
69
LFM2.5-1.2B-Instruct
80
DeepSeek-R1
60.4
LFM2.5-1.2B-Instruct
60.7
DeepSeek-R1
52.5
LFM2.5-1.2B-Instruct
37
DeepSeek-R1 is ahead overall, 43 to 30. The biggest single separator in this matchup is LongBench v2, where the scores are 58 and 34.
DeepSeek-R1 has the edge for knowledge tasks in this comparison, averaging 37.9 versus 26. Inside this category, MMLU is the benchmark that creates the most daylight between them.
DeepSeek-R1 has the edge for coding in this comparison, averaging 21.5 versus 7.2. Inside this category, HumanEval is the benchmark that creates the most daylight between them.
DeepSeek-R1 has the edge for math in this comparison, averaging 52.5 versus 37. Inside this category, AIME 2023 is the benchmark that creates the most daylight between them.
DeepSeek-R1 has the edge for reasoning in this comparison, averaging 50.9 versus 32.1. Inside this category, LongBench v2 is the benchmark that creates the most daylight between them.
DeepSeek-R1 has the edge for agentic tasks in this comparison, averaging 44.5 versus 25.7. Inside this category, Terminal-Bench 2.0 is the benchmark that creates the most daylight between them.
DeepSeek-R1 has the edge for multimodal and grounded tasks in this comparison, averaging 47.5 versus 32.4. Inside this category, MMMU-Pro is the benchmark that creates the most daylight between them.
LFM2.5-1.2B-Instruct has the edge for instruction following in this comparison, averaging 80 versus 69. Inside this category, IFEval is the benchmark that creates the most daylight between them.
LFM2.5-1.2B-Instruct has the edge for multilingual tasks in this comparison, averaging 60.7 versus 60.4. 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.