Skip to content

Benchmark

SurveyReview v1.0 · Criticalness · MAE

Lower scores are better.

Models and systems measured
7
Scores recorded
7
Sources
1
IQ weight
No weight published

Score distribution

1.1
Mean
1
Median
0.218
Std dev
0.88
10th percentile
1.2
90th percentile
0.82
Min
0.821.5
The band covers the middle 80% of scores. The line marks the median.

Leaderboard

#Model or systemScore
1Claude Opus 4.5SurveyReview v1.0 published leaderboard · text survey-paper evaluation · exact candidate snapshot, reasoning effort, prompt, run count, and harness are not specified in the summary table· Reported by SurveyReview authors / Z.ai· via BenchBoard docs0.88
2GPT 5.2SurveyReview v1.0 published leaderboard · text survey-paper evaluation · exact candidate snapshot, reasoning effort, prompt, run count, and harness are not specified in the summary table· Reported by SurveyReview authors / Z.ai· via BenchBoard docs0.97
3Gemini 3 Pro PreviewSurveyReview v1.0 published leaderboard · text survey-paper evaluation · exact candidate snapshot, reasoning effort, prompt, run count, and harness are not specified in the summary table· Reported by SurveyReview authors / Z.ai· via BenchBoard docs1
4DeepSeek V3.2SurveyReview v1.0 published leaderboard · text survey-paper evaluation · exact candidate snapshot, reasoning effort, prompt, run count, and harness are not specified in the summary table· Reported by SurveyReview authors / Z.ai· via BenchBoard docs1.2

Not ranked

The source marks these results as descriptive, not capability scores, so we list them by name instead of ranking them.

Model or systemScore
GLM-4.7 (SurveyReview v1.0 label)SurveyReview v1.0 published leaderboard · text survey-paper evaluation · exact candidate snapshot, reasoning effort, prompt, run count, and harness are not specified in the summary table· Reported by SurveyReview authors / Z.ai· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.1.2
Qwen3-32B (SurveyReview v1.0 label)SurveyReview v1.0 published leaderboard · text survey-paper evaluation · exact candidate snapshot, reasoning effort, prompt, run count, and harness are not specified in the summary table· Reported by SurveyReview authors / Z.ai· via BenchBoard docs1.5
SurveyReviewer (SurveyReview v1.0 label)SurveyReview v1.0 published leaderboard · text survey-paper evaluation · exact candidate snapshot, reasoning effort, prompt, run count, and harness are not specified in the summary table· Reported by SurveyReview authors / Z.ai· via BenchBoard docs· Reported by the model's developer. The source relaying it did not verify it.0.82

Data credits: BenchBoard by SYuan03 (github.com/SYuan03/benchboard). MIT covers its code and data organization; no results licence is inferred. Numeric relays retain the original stating-source links.