New: connect Claude & other AIs to GenAIList over MCP — research the catalog and contribute to the shared knowledge base. Learn how →
INCLUDE benchmark
AI model leaderboard for the INCLUDE benchmark. Compare how large language models score on INCLUDE, see the full ranking, and understand what this AI benchmark measures. Qwen3-235B-A22B currently leads with 78.7. Multilingual knowledge benchmark — covers 65 languages and tests world-knowledge breadth.
Leaderboard
| # | Model | Organization | Score | Variant | Source |
|---|---|---|---|---|---|
| #1 | Qwen3-235B-A22B | Qwen | 78.7 | thinking | official ↗ |
| #2 | Qwen3-235B-A22B | Qwen | 75.6 | non-thinking | official ↗ |
| #3 | Qwen3-32B | Qwen | 73.7 | thinking | official ↗ |
| #4 | Qwen3-30B-A3B | Qwen | 71.9 | thinking | official ↗ |
| #5 | Qwen3-14B | Qwen | 71.7 | thinking | official ↗ |
| #6 | Qwen3-32B | Qwen | 70.9 | non-thinking | official ↗ |
| #7 | Qwen3-30B-A3B | Qwen | 67.8 | non-thinking | official ↗ |
| #8 | Qwen3-14B | Qwen | 67.8 | non-thinking | official ↗ |
| #9 | Qwen3-8B | Qwen | 67.8 | thinking | official ↗ |
| #10 | Granite-4.0-H-Small | IBM | 63.97 | 5-shot | official ↗ |
| #11 | Qwen3-8B | Qwen | 62.5 | non-thinking | official ↗ |
| #12 | Qwen3-4B | Qwen | 61.8 | thinking | official ↗ |
| #13 | Qwen3-4B | Qwen | 53.8 | non-thinking | official ↗ |
| #14 | Granite-4.0-H-Tiny | IBM | 53.12 | 5-shot | official ↗ |
| #15 | Qwen3-1.7B | Qwen | 51.8 | thinking | official ↗ |
| #16 | Granite-4.0-H-Micro | IBM | 50.51 | 5-shot | official ↗ |
| #17 | Qwen3-1.7B | Qwen | 42.6 | non-thinking | official ↗ |
| #18 | Qwen3-0.6B | Qwen | 35.9 | thinking | official ↗ |
| #19 | Qwen3-0.6B | Qwen | 34.4 | non-thinking | official ↗ |
Frequently asked questions about INCLUDE
What is the INCLUDE benchmark?
Multilingual knowledge benchmark — covers 65 languages and tests world-knowledge breadth.
How is the INCLUDE benchmark scored?
INCLUDE is scored using the accuracy metric, where a higher score is better. GenAIList aggregates reported scores from model providers and papers into a single ranked leaderboard.
Which AI model scores highest on INCLUDE?
As of the latest reported scores on GenAIList, Qwen3-235B-A22B achieves the highest result on INCLUDE with a score of 78.7.
Is a higher INCLUDE score better?
Yes. On INCLUDE a higher score indicates better performance, so models near the top of the leaderboard are the strongest.