Model
Qwen 3.5
Alibaba · open weights
Knowledge & Reasoning
Knowledge & Reasoning| Benchmark | Score | Metric | Harness | Source | Confidence | Link |
|---|---|---|---|---|---|---|
| GPQA Diamond | 88.4 | accuracy | N/A | Self-reported | low | artificialanalysis.ai |
| HLE | 28.7 | accuracy | N/A | Self-reported | low | lastexam.ai |
| LMArena | 1450.0 | elo | N/A | Self-reported | low | lmarena.ai |
| MMLU-Pro | 87.8 | accuracy | N/A | Self-reported | low | artificialanalysis.ai |
Coding
Coding| Benchmark | Score | Metric | Harness | Source | Confidence | Link |
|---|---|---|---|---|---|---|
| LiveCodeBench | 83.6 | pass@1 | N/A | Self-reported | low | livecodebench.github.io |
| SWE-bench Verified | 76.4 | resolved_rate | N/A | Self-reported | low | swebench.com |
| Terminal-Bench 2.0 | 52.5 | accuracy | N/A | Self-reported | low | tbench.ai |
Agentic / Tool Use
Agentic / Tool Use| Benchmark | Score | Metric | Harness | Source | Confidence | Link |
|---|---|---|---|---|---|---|
| BrowseComp | 78.6 | accuracy | N/A | Self-reported | low | openai.com |
| OSWorld | 62.2 | success_rate | N/A | Self-reported | low | os-world.github.io |
| τ²-bench | 86.7 | pass^1 | N/A | Self-reported | low | github.com |
Multimodal
Multimodal| Benchmark | Score | Metric | Harness | Source | Confidence | Link |
|---|---|---|---|---|---|---|
| EditJudgeBias | 3.5 | rubric_ratings_(1-10)_and_pairwise_preferences | N/A | Independent | medium | arxiv.org |
| MMMU-Pro | 79.0 | accuracy | N/A | Self-reported | low | artificialanalysis.ai |
Instruction Following
Instruction Following| Benchmark | Score | Metric | Harness | Source | Confidence | Link |
|---|---|---|---|---|---|---|
| IFEval | 92.6 | prompt_strict_accuracy | N/A | Self-reported | low | artificialanalysis.ai |
Reproduced by the community
Loading community runs…