Agents · Benchmark profile
An interactive real-world agent benchmark grounded in practical consumer-service tasks such as delivery, in-store consumption, and online travel workflows.
Data verified 27 Jul 2026 · Methodology 1.8.0
Benchmark score on VITA-Bench
Qwen3.7-Max leads at 47.9%, followed by Qwen3.7-Plus (45.6%) and Qwen3.6 Plus (44.3%).
10 ranked models · higher is better · labels show rank and score
| Model | Rank | Provider | Score |
|---|---|---|---|
| Qwen3.7-Max | #1 | Alibaba Cloud | 47.9% |
| Qwen3.7-Plus | #2 | Alibaba Cloud | 45.6% |
| Qwen3.6 Plus | #3 | Alibaba Cloud | 44.3% |
| Qwen3.5 397B | #4 | Alibaba Cloud | 43.7% |
| Agents-A1 | #5 | InternScience | 38.8% |
| Qwen3.6-35B-A3B | #6 | Alibaba Cloud | 35.6% |
| Claude Opus 4.5 | #7 | Anthropic | 23.3% |
| DeepSeek V3.2 | #8 | DeepSeek | 18.5% |
| Claude Sonnet 4.5 | #9 | Anthropic | 17% |
| GLM-4.7 | #10 | Z.AI | 15.5% |
One best score per model · higher is better
| Rank | Model | Provider | License | Evidence use | Score |
|---|---|---|---|---|---|
| #1 | Qwen3.7-Max qwen3.7-max | Alibaba Cloud | closed | Reference only | 47.9% |
| #2 | Qwen3.7-Plus qwen3.7-plus | Alibaba Cloud | closed | Reference only | 45.6% |
| #3 | Qwen3.6 Plus qwen3-6-plus | Alibaba Cloud | closed | Reference only | 44.3% |
| #4 | Qwen3.5 397B qwen3.5-397b-a17b | Alibaba Cloud | open | Estimated reference | 43.7% |
| #5 | Agents-A1 agents-a1 | InternScience | open | Estimated reference | 38.8% |
| #6 | Qwen3.6-35B-A3B qwen3-6-35b-a3b | Alibaba Cloud | open | Estimated reference | 35.6% |
| #7 | Claude Opus 4.5 claude-opus-4-5 | Anthropic | closed | Reference only | 23.3% |
| #8 | DeepSeek V3.2 deepseek-v3.2 | DeepSeek | open | Reference only | 18.5% |
| #9 | Claude Sonnet 4.5 claude-sonnet-4-5 | Anthropic | closed | Estimated reference | 17% |
| #10 | GLM-4.7 glm-4.7 | Z.AI | open | Reference only | 15.5% |
The top of this snapshot is led by Qwen3.7-Max at 47.9%; third place is 3.6 points behind. The top-10 spread is 32.4 points.
About VITA-Bench
An interactive real-world agent benchmark grounded in practical consumer-service tasks such as delivery, in-store consumption, and online travel workflows. Results stay tied to the exact model variant and evaluation system. Multiple systems for the same model use the best published score on this page; overall Lumina scoring uses the median of ranking-eligible rows.
Open benchmark source ↗FAQ
An interactive real-world agent benchmark grounded in practical consumer-service tasks such as delivery, in-store consumption, and online travel workflows.
Qwen3.7-Max by Alibaba Cloud currently leads with 47.9%.
10 models in the LuminaBench cohort have a qualifying score on this benchmark.
No. This benchmark is display-only and does not enter the overall Lumina composite.
Related
