LongBench v2
Multilingual
text
+
+
+
+
About
LongBench v2 is an advanced long-context evaluation benchmark designed to assess LLMs' ability to handle complex problems requiring deep understanding and reasoning across real-world multitasks. This benchmark tests models' sustained attention, information integration, and reasoning capabilities when processing extended contexts, measuring performance on tasks that demand comprehensive understanding of lengthy documents.
+
+
+
+
Evaluation Stats
Total Models1
Organizations1
Verified Results0
Self-Reported1
+
+
+
+
Benchmark Details
Max Score1
Language
en
+
+
+
+
Performance Overview
Score distribution and top performers
Score Distribution
1 models
Top Score
48.7%
Average Score
48.7%
High Performers (80%+)
0Top Organizations
#1DeepSeek
1 model
48.7%
+
+
+
+
Leaderboard
1 models ranked by performance on LongBench v2
| License | Links | ||||
|---|---|---|---|---|---|
| Dec 25, 2024 | MIT + Model License (Commercial use allowed) | 48.7% |
