InternVL3_5-2B vs Qwen3.5 2B

OpenGVLab

Wins 1 of 2 areas

Images and charts

Alibaba · released Mar 2, 2026

Wins 1 of 2 areas

Following instructions

The two are evenly matched.InternVL3_5-2B is better at images and charts; Qwen3.5 2B at following instructions.

Scores updated Sep 19, 2026 · 29 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 video32InternVL3_5-2B3 of 5 tests
Following instructionsDoing exactly what it is asked01Qwen3.5 2B1 of 1 test

Following instructions rests on a single test.

The biggest differences

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

Where InternVL3_5-2B pulls ahead

Math problems shown in pictures and chartsMathVista+8.1points aheadInternVL3_5-2B56.8Qwen3.5 2B48.7
Visual perception tasks people solve at a glanceBLINK+3.4points aheadInternVL3_5-2B52Qwen3.5 2B48.6
Questions that truly need the image to answerMMStar+2.6points aheadInternVL3_5-2B57.7Qwen3.5 2B55.1

Where Qwen3.5 2B pulls ahead

Harder college exam questions with imagesMMMU-Pro+21.7points aheadQwen3.5 2B43InternVL3_5-2B21.3
Follows unfamiliar, precisely checkable instructionsIFBench+7.1points aheadQwen3.5 2B31.5InternVL3_5-2B24.4
Reads text in imagesOCRBench+2.7points aheadQwen3.5 2B86.1InternVL3_5-2B83.4

Every test, side by side

All 29 tests both models report. The winning score is in its model's colour; marks a score checked independently.

Images and chartsInternVL3_5-2B

MMMU-ProQwen3.5 2B by 21.721.343+21.7
MathVistaInternVL3_5-2B by 8.156.848.7+8.1
BLINKInternVL3_5-2B by 3.45248.6+3.4
OCRBenchQwen3.5 2B by 2.783.486.1+2.7
MMStarInternVL3_5-2B by 2.657.755.1+2.6

Full images and charts ranking

Following instructionsQwen3.5 2B

IFBenchQwen3.5 2B by 7.124.431.5+7.1

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.

IFEvalQwen3.5 2B by 41.232.473.6+41.2
HallusionBenchQwen3.5 2B by 17.947.665.5+17.9
ScreenSpot-v2 MobileInternVL3_5-2B by 16.586.269.7+16.5
ScreenSpot-v2 DesktopInternVL3_5-2B by 16.179.963.8+16.1
ScreenSpot-v2 WebInternVL3_5-2B by 1479.965.9+14
MM-IFEvalQwen3.5 2B by 8.347.155.4+8.3
MMMU (val) (Pass@1)InternVL3_5-2B by 7.95244.1+7.9
SimpleVQAQwen3.5 2B by 4.730.535.2+4.7
DocVQA-valQwen3.5 2B by 4.288.492.6+4.2
InfographicVQA (val)Qwen3.5 2B by 4.269.373.5+4.2
RealWorldQAQwen3.5 2B by 3.561.665.1+3.5
ChartQA TestInternVL3_5-2B by 3.381.778.4+3.3
MMBenchInternVL3_5-2B by 3.176.273.1+3.1
MMEQwen3.5 2B by 2.673.676.2+2.6
OCRBench v2_enQwen3.5 2B by 2.645.548.1+2.6
VLM Judge ScoreQwen3.5 2B by 287.789.7+2
MMMBInternVL3_5-2B by 1.876.374.5+1.8
Multilingual MMBenchInternVL3_5-2B by 170.969.9+1
TextVQA-valtie76.677.3tie
POPEtie8888.6tie
OCRBench v1tie83.984.4tie
Multilingual MMMBtie76.375.9tie
SEED-Bench (image)tie75.475.8tie

Questions people ask

Which is better, InternVL3_5-2B or Qwen3.5 2B?

InternVL3_5-2B and Qwen3.5 2B each win one of the two areas where both have results. InternVL3_5-2B wins images and charts; Qwen3.5 2B wins following instructions.

How do you compare the two?

We use the 29 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 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.