Skip to content
VECTOR WIREAI INTELLIGENCE
UTC

Gemma 4 E2B vs LFM2-VL-3B

Google · released

Wins 1 of 2 areas

Following instructions

Liquid AI · released

Wins 1 of 2 areas

Images and charts

The two are evenly matched.Gemma 4 E2B is better at following instructions; LFM2-VL-3B at images and charts.

Scores updated · 32 tests both models report · How we compare

Where each one wins

Tests won in each of the two 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.

Following instructions rests on a single test.

The biggest differences

The tests each model wins by the widest margin, up to three each. Scores are out of 100.

Where LFM2-VL-3B pulls ahead

  • Math problems shown in pictures and chartsMathVista+24.3points aheadLFM2-VL-3B62.1Gemma 4 E2B37.8
  • Questions that truly need the image to answerMMStar+12.4points aheadLFM2-VL-3B57.7Gemma 4 E2B45.3
  • Reads text in imagesOCRBench+10.3points aheadLFM2-VL-3B82.2Gemma 4 E2B71.9

Where Gemma 4 E2B pulls ahead

  • Follows unfamiliar, precisely checkable instructionsIFBench+17.2points aheadGemma 4 E2B38LFM2-VL-3B20.8
  • Harder college exam questions with imagesMMMU-Pro+15.9points aheadGemma 4 E2B44.6LFM2-VL-3B28.7

Every test, side by side

All 32 tests both models report. The winning score is in its model's colour; marks a score checked independently.

Images and chartsLFM2-VL-3B

Full images and charts ranking

Following instructionsGemma 4 E2B
  • IFBenchGemma 4 E2B by 17.23820.8+17.2

Full following instructions ranking

Other results25 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.

Questions people ask

Which is better, Gemma 4 E2B or LFM2-VL-3B?

Gemma 4 E2B and LFM2-VL-3B each win one of the two areas where both have results. Gemma 4 E2B wins following instructions; LFM2-VL-3B wins images and charts.

How do you compare the two?

We use the 32 benchmark tests both models have published scores on. The verdict counts the 7 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 25 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.