Gemma 4 E4B vs LFM2.5-VL-3B

Google · released Apr 3, 2026

Wins 1 of 2 areas

Following instructions

Liquid AI · released Aug 11, 2026

Wins 1 of 2 areas

Images and charts

The two are evenly matched.Gemma 4 E4B is better at following instructions; LFM2.5-VL-3B at images and charts.

Scores updated Sep 19, 2026 · 28 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.

Images and chartsUnderstanding pictures, charts and video13LFM2.5-VL-3B3 of 4 tests
Following instructionsDoing exactly what it is asked10Gemma 4 E4B1 of 1 test

Following instructions rests on a single test.

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.

Gemma 4 E4B$0.02 to read · $0.10 to write$0.12Price from Artificial Analysis

LFM2.5-VL-3BNo current price is tracked.

The biggest differences

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

Where LFM2.5-VL-3B pulls ahead

Math problems shown in pictures and chartsMathVista+23.3points aheadLFM2.5-VL-3B68.5Gemma 4 E4B45.2
Questions that truly need the image to answerMMStar+10.4points aheadLFM2.5-VL-3B63.3Gemma 4 E4B52.9
Visual perception tasks people solve at a glanceBLINK+9.3points aheadLFM2.5-VL-3B61.5Gemma 4 E4B52.2

Where Gemma 4 E4B pulls ahead

Harder college exam questions with imagesMMMU-Pro+20.9points aheadGemma 4 E4B51.4LFM2.5-VL-3B30.5
Follows unfamiliar, precisely checkable instructionsIFBench+18.4points aheadGemma 4 E4B44.2LFM2.5-VL-3B25.8

Every test, side by side

All 28 tests both models report. The winning score is in its model's colour; marks a score checked independently.

Images and chartsLFM2.5-VL-3B

MathVistaLFM2.5-VL-3B by 23.345.268.5+23.3
MMMU-ProGemma 4 E4B by 20.951.430.5+20.9
MMStarLFM2.5-VL-3B by 10.452.963.3+10.4
BLINKLFM2.5-VL-3B by 9.352.261.5+9.3

Full images and charts ranking

Following instructionsGemma 4 E4B

IFBenchGemma 4 E4B by 18.444.225.8+18.4

Full following instructions ranking

Other results23 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.

ChartQA TestLFM2.5-VL-3B by 39.242.181.3+39.2
ScreenSpot-v2 WebLFM2.5-VL-3B by 34.647.682.2+34.6
ScreenSpot-v2 DesktopLFM2.5-VL-3B by 32.945.878.7+32.9
ScreenSpot-v2 MobileLFM2.5-VL-3B by 20.960.381.2+20.9
TextVQA-valLFM2.5-VL-3B by 15.36984.3+15.3
OCRBench v1LFM2.5-VL-3B by 10.773.584.2+10.7
MMBenchLFM2.5-VL-3B by 9.471.681+9.4
InfographicVQA (val)LFM2.5-VL-3B by 9.360.970.2+9.3
RealWorldQALFM2.5-VL-3B by 8.864.373.1+8.8
Multilingual MMBenchLFM2.5-VL-3B by 8.371.279.5+8.3
MM-IFEvalGemma 4 E4B by 7.668.260.6+7.6
BFCLv4Gemma 4 E4B by 7.54032.5+7.5
IFEvalGemma 4 E4B by 5.687.982.3+5.6
MMELFM2.5-VL-3B by 5.567.673.1+5.5
SimpleVQALFM2.5-VL-3B by 530.435.4+5
DocVQA-valLFM2.5-VL-3B by 3.787.491.1+3.7
HallusionBenchGemma 4 E4B by 2.649.847.2+2.6
MMMBLFM2.5-VL-3B by 2.680.483+2.6
Multilingual MMMBLFM2.5-VL-3B by 2.680.483+2.6
SEED-Bench (image)LFM2.5-VL-3B by 2.475.377.7+2.4
POPELFM2.5-VL-3B by 1.886.988.7+1.8
OCRBench v2_enGemma 4 E4B by 1.348.847.5+1.3
MMMU (val) (Pass@1)tie49.348.4tie

Questions people ask

Which is better, Gemma 4 E4B or LFM2.5-VL-3B?

Gemma 4 E4B and LFM2.5-VL-3B each win one of the two areas where both have results. Gemma 4 E4B wins following instructions; LFM2.5-VL-3B wins images and charts.

How do you compare the two?

We use the 28 benchmark tests both models have published scores on. The verdict counts the 5 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 23 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.