LLM Reference
active

GAOKAO-MM

Metric: Accuracy (higher is better)Introduced: 2024

Multimodal extension of GAOKAO with text+image questions from Chinese college entrance exams, evaluating vision-language model capabilities.

Models ranked

7

tracked on this benchmark

Score band

48.1 – 14.2

best → lowest tracked

Snapshot trend

need ≥2 snapshots

Leaderboard

Tracked models ranked by Accuracy (higher is better).

Compare candidates
#Model variant and provenanceScore
1
GPT-4 Vision Preview
Version: zero-shotHarness: Not recordedEvaluator: Not recordedObserved: Feb 23, 2024Confidence: Not recordedSource
48.1
2
Qwen-VL-Plus
Version: zero-shotHarness: Not recordedEvaluator: Not recordedObserved: Feb 23, 2024Confidence: Not recordedSource
41.2
3
Gemini 1.0 Pro Vision
Version: zero-shotHarness: Not recordedEvaluator: Not recordedObserved: Feb 23, 2024Confidence: Not recordedSource
35.1
4
InternLM XComposer2 VL 7B
Version: zero-shotHarness: Not recordedEvaluator: Not recordedObserved: Feb 23, 2024Confidence: Not recordedSource
33.2
5
LLaVA 1.5 13B
Version: zero-shotHarness: Not recordedEvaluator: Not recordedObserved: Feb 23, 2024Confidence: Not recordedSource
18.0
6
LLaVA 1.5 7B
Version: zero-shotHarness: Not recordedEvaluator: Not recordedObserved: Feb 23, 2024Confidence: Not recordedSource
17.9
7
Qwen-VL
Version: zero-shotHarness: Not recordedEvaluator: Not recordedObserved: Feb 23, 2024Confidence: Not recordedSource
14.2

How to read this benchmark

This benchmark scores models where higher is better. Use scores for directional filtering and shortlisting, not universal quality ranking; then validate pricing, context window, provider availability, and fit for your workload.

Trust this score when

  • There is a fresh timestamped snapshot (or multiple snapshots) for this benchmark.
  • The model list covers the same version family you can actually deploy today.
  • Top candidates overlap with your required routing and feature requirements.

Be cautious when

  • There is only one benchmark snapshot or the dataset appears stale.
  • The benchmark metric direction is opposite of your decision objective.
  • The score difference between options is narrow and likely within implementation variance.

Last reviewed: May 1, 2026