Gemma 4 E4B vs LFM2-VL-3B

Google · released Apr 3, 2026

Wins 1 of 2 areas

Following instructions

Liquid AI · released Oct 22, 2025

Wins 0 of 2 areas

—

Gemma 4 E4B wins more areas, narrowly.

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.

Following instructionsDoing exactly what it is asked10Gemma 4 E4B1 of 1 test
Images and chartsUnderstanding pictures, charts and video22Even2 each

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-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 Gemma 4 E4B pulls ahead

Follows unfamiliar, precisely checkable instructionsIFBench+23.4points aheadGemma 4 E4B44.2LFM2-VL-3B20.8
Harder college exam questions with imagesMMMU-Pro+22.7points aheadGemma 4 E4B51.4LFM2-VL-3B28.7
Visual perception tasks people solve at a glanceBLINK+2points aheadGemma 4 E4B52.2LFM2-VL-3B50.2

Where LFM2-VL-3B pulls ahead

Math problems shown in pictures and chartsMathVista+16.9points aheadLFM2-VL-3B62.1Gemma 4 E4B45.2
Questions that truly need the image to answerMMStar+4.8points aheadLFM2-VL-3B57.7Gemma 4 E4B52.9

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 chartsEven

MMMU-ProGemma 4 E4B by 22.751.428.7+22.7
MathVistaLFM2-VL-3B by 16.945.262.1+16.9
MMStarLFM2-VL-3B by 4.852.957.7+4.8
BLINKGemma 4 E4B by 252.250.2+2

Full images and charts ranking

Following instructionsGemma 4 E4B

IFBenchGemma 4 E4B by 23.444.220.8+23.4

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.

ScreenSpot-v2 MobileGemma 4 E4B by 52.760.37.6+52.7
ScreenSpot-v2 WebGemma 4 E4B by 45.147.62.5+45.1
ScreenSpot-v2 DesktopGemma 4 E4B by 39.845.86+39.8
ChartQA TestLFM2-VL-3B by 38.342.180.4+38.3
BFCLv4Gemma 4 E4B by 19.54020.5+19.5
MM-IFEvalGemma 4 E4B by 16.868.251.4+16.8
IFEvalGemma 4 E4B by 1587.972.9+15
TextVQA-valLFM2-VL-3B by 146983+14
MMBenchLFM2-VL-3B by 8.471.680+8.4
OCRBench v1LFM2-VL-3B by 8.273.581.7+8.2
InfographicVQA (val)LFM2-VL-3B by 6.960.967.8+6.9
RealWorldQALFM2-VL-3B by 6.864.371.1+6.8
MMELFM2-VL-3B by 5.467.673+5.4
Multilingual MMBenchLFM2-VL-3B by 5.171.276.3+5.1
OCRBench v2_enGemma 4 E4B by 4.948.843.9+4.9
MMMU (val) (Pass@1)Gemma 4 E4B by 3.749.345.6+3.7
HallusionBenchGemma 4 E4B by 3.449.846.4+3.4
SimpleVQALFM2-VL-3B by 2.630.433+2.6
DocVQA-valLFM2-VL-3B by 2.487.489.8+2.4
POPELFM2-VL-3B by 2.386.989.2+2.3
MMMBLFM2-VL-3B by 1.580.481.9+1.5
Multilingual MMMBLFM2-VL-3B by 1.580.481.9+1.5
SEED-Bench (image)LFM2-VL-3B by 1.375.376.6+1.3

Questions people ask

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

Gemma 4 E4B wins one of the two areas where both have results: following instructions. LFM2-VL-3B wins none. They are level on 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.