Gemma 4 E2B vs LFM2.5-2.6B
Wins 1 of 6 areas
Long documents
Wins 3 of 6 areas
Agents · Reasoning · Following instructions
LFM2.5-2.6B wins more areas, narrowly.Gemma 4 E2B is better at long documents.
Scores updated · 23 tests both models report · How we compare
Where each one wins
Tests won in each of the six 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.
- ReasoningHard problems that need careful thinking02LFM2.5-2.6B2 of 3 tests · 1 tie
- AgentsCarrying out multi-step tasks on its own01LFM2.5-2.6B1 of 2 tests · 1 tie
- Following instructionsDoing exactly what it is asked01LFM2.5-2.6B1 of 1 test
- Long documentsFinding answers in very long texts10Gemma 4 E2B1 of 1 test
- CodingWriting and fixing software11Even1 each
- FactsGetting facts right instead of making them up11Even1 each
Long documents and following instructions rest on a single test each.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where LFM2.5-2.6B pulls ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+21.2points ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+16.4points ahead
- Recent programming contest problemsLiveCodeBench v6+15.4points ahead
Where Gemma 4 E2B pulls ahead
- Reasons across sets of long documentsAA-LCR+10.6points ahead
- Code for real scientific research problemsSciCode+6.5points ahead
- Answers hard knowledge questions correctlyAA-Omniscience · Accuracy+2.2points ahead
Every test, side by side
All 23 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingEven
- LiveCodeBench v6LFM2.5-2.6B by 15.44459.4+15.4
- SciCodeGemma 4 E2B by 6.520.914.4+6.5
ReasoningLFM2.5-2.6B
- GPQA DiamondLFM2.5-2.6B by 12.543.355.8+12.5
- Humanity's Last ExamLFM2.5-2.6B by 1.44.86.2+1.4
- CritPttie00tie
FactsEven
- AA-Omniscience · Non-hallucinationLFM2.5-2.6B by 16.467.684+16.4
- AA-Omniscience · AccuracyGemma 4 E2B by 2.26.64.3+2.2
Following instructionsLFM2.5-2.6B
- IFBenchLFM2.5-2.6B by 21.23859.2+21.2
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.
- MATH500 (Pass@1)Gemma 4 E2B by 58.6645.4+58.6
- AIME25 no toolsLFM2.5-2.6B by 25.626.351.9+25.6
- PinchBenchLFM2.5-2.6B by 2444.268.2+24
- BFCLv4LFM2.5-2.6B by 23.733.256.9+23.7
- BrowseComp+ (OpenClaw)LFM2.5-2.6B by 18.68.326.9+18.6
- AA-OmniscienceLFM2.5-2.6B by 12.7-23.6-10.9+12.7
- Claw-Eval AvgLFM2.5-2.6B by 9.853.162.9+9.8
- LiveCodeBenchV6 no toolsLFM2.5-2.6B by 4.554.959.4+4.5
- τ³-Bench BankingLFM2.5-2.6B by 2.33.45.7+2.3
- AA Agentic Indextie1.52.4tie
- AA Intelligencetie7.88.4tie
- Artificial Analysis Coding Indextie7.27.7tie
Questions people ask
Which is better, Gemma 4 E2B or LFM2.5-2.6B?
LFM2.5-2.6B wins three of the six areas where both have results: agents, reasoning and following instructions. Gemma 4 E2B wins long documents. They are level on coding and facts.
Which is better for coding?
Neither. They win 1 coding test each of the 2 both models report.
How do you compare the two?
We use the 23 benchmark tests both models have published scores on. The verdict counts the 11 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.