LFM2.5-2.6B vs LFM3-1.7B
Wins 0 of 4 areas
—
Wins 4 of 4 areas
Coding · Agents · Reasoning · Following instructions
LFM3-1.7B is the stronger all-rounder.
Scores updated · 18 tests both models report · How we compare
Where each one wins
Tests won in each of the four areas where both have results. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
Agents and following instructions rest on a single test each.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where LFM3-1.7B pulls ahead
- Multi-step tasks using many real software toolsMCP Atlas+16.1points ahead
- Recent programming contest problemsLiveCodeBench v6+14.1points ahead
- Code for real scientific research problemsSciCode+14.1points ahead
Where LFM2.5-2.6B pulls ahead
No clear win on a test scored out of 100.
Every test, side by side
All 18 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingLFM3-1.7B
- LiveCodeBench v6LFM3-1.7B by 14.159.473.5+14.1
- SciCodeLFM3-1.7B by 14.114.428.5+14.1
ReasoningLFM3-1.7B
- GPQA DiamondLFM3-1.7B by 7.655.863.4+7.6
- Humanity's Last ExamLFM3-1.7B by 16.27.2+1
Following instructionsLFM3-1.7B
- IFBenchLFM3-1.7B by 13.859.273+13.8
Other results12 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- WebArena-VerifiedLFM3-1.7B by 35.4338.4+35.4
- EnterpriseOps-GymLFM3-1.7B by 20.37.928.2+20.3
- SWE-bench Multilingual (OpenCode)LFM3-1.7B by 15823+15
- BFCLv4LFM3-1.7B by 14.556.971.4+14.5
- Claw-Eval General (EN), pass^3LFM3-1.7B by 13.937.551.4+13.9
- Terminal-Bench 2.1 (OpenCode)LFM3-1.7B by 12.9012.9+12.9
- LiveCodeBench Pro 25Q2LFM3-1.7B by 12.519.231.7+12.5
- τ³-bench (4 domains), averageLFM3-1.7B by 11.256.868+11.2
- BrowseComp-Plus (OpenClaw subset)LFM3-1.7B by 10.726.337+10.7
- Claw-Eval General (EN), averageLFM3-1.7B by 662.668.6+6
- HLE (with tools)LFM3-1.7B by 4.39.613.9+4.3
- IFEvalLFM3-1.7B by 1.395.496.7+1.3
Questions people ask
Which is better, LFM2.5-2.6B or LFM3-1.7B?
LFM3-1.7B wins all four areas where both have results: coding, agents, reasoning and following instructions. LFM2.5-2.6B wins none.
Which is better for coding?
LFM3-1.7B. It wins 2 of the 2 coding tests both models report; LFM2.5-2.6B wins none.
How do you compare the two?
We use the 18 benchmark tests both models have published scores on. The verdict counts the 6 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 12 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.