New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
LongBench v2 benchmark
AI model leaderboard for the LongBench v2 benchmark. Compare how large language models score on LongBench v2, see the full ranking, and understand what this AI benchmark measures. MiniMax-M1-80k currently leads with 61.5%. Long-context understanding benchmark across 8 task types and contexts up to 2M tokens.
Leaderboard
| # | Model | Organization | Score | Variant | Source |
|---|---|---|---|---|---|
| #1 | MiniMax-M1-80k | MiniMax | 61.5% | official ↗ | |
| #2 | MiniMax-M1-40k | MiniMax | 61% | official ↗ | |
| #3 | Kimi Linear | Moonshot | 35% | long-context-128k | official ↗ |
Frequently asked questions about LongBench v2
What is the LongBench v2 benchmark?
Long-context understanding benchmark across 8 task types and contexts up to 2M tokens.
How is the LongBench v2 benchmark scored?
LongBench v2 is scored using the accuracy (%) metric, where a higher score is better. The maximum achievable score is 100.000. GenAIList aggregates reported scores from model providers and papers into a single ranked leaderboard.
Which AI model scores highest on LongBench v2?
As of the latest reported scores on GenAIList, MiniMax-M1-80k achieves the highest result on LongBench v2 with a score of 61.5%.
Is a higher LongBench v2 score better?
Yes. On LongBench v2 a higher score indicates better performance, so models near the top of the leaderboard are the strongest.