Skip to content
VECTOR WIREAI INTELLIGENCE
UTC

InternVL2.5-78B vs Phi 4 Multimodal Instruct

OpenGVLab

Wins 1 of 1 area

Images and charts

Microsoft · released

Wins 0 of 1 area

—

InternVL2.5-78B is the stronger all-rounder.

Scores updated · 7 tests both models report · How we compare

Where each one wins

Tests won in each of the one area 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.

The biggest differences

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

Where InternVL2.5-78B pulls ahead

  • Harder college exam questions with imagesMMMU-Pro+32.8points aheadInternVL2.5-78B47.3Phi 4 Multimodal Instruct14.5
  • College exam questions with charts, maps and diagramsMMMU+11.4points aheadInternVL2.5-78B66.5Phi 4 Multimodal Instruct55.1
  • Math problems shown in pictures and chartsMathVista+6points aheadInternVL2.5-78B68.4Phi 4 Multimodal Instruct62.4

Where Phi 4 Multimodal Instruct pulls ahead

No clear win on a test scored out of 100.

Every test, side by side

All 7 tests both models report. The winning score is in its model's colour; marks a score checked independently.

Images and chartsInternVL2.5-78B
  • MMMU-ProInternVL2.5-78B by 32.847.314.5+32.8
  • MMMUInternVL2.5-78B by 11.466.555.1+11.4
  • MathVistaInternVL2.5-78B by 668.462.4+6
  • OCRBenchtie84.784.4tie

Full images and charts ranking

Other results3 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.

  • ChartQAInternVL2.5-78B by 10.191.581.4+10.1
  • AI2DInternVL2.5-78B by 4.586.882.3+4.5
  • DocVQAInternVL2.5-78B by 2.996.193.2+2.9

Questions people ask

Which is better, InternVL2.5-78B or Phi 4 Multimodal Instruct?

InternVL2.5-78B wins the one area where both have results: images and charts. Phi 4 Multimodal Instruct wins none.

How do you compare the two?

We use the 7 benchmark tests both models have published scores on. The verdict counts the 4 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 3 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.