Long Context · Benchmark profile
A long-context reasoning benchmark built around demanding real-world tasks.
Data verified 15 July 2026 · Methodology 1.2.0
Benchmark score on LongBench v2
Qwen3.7-Plus leads at 91.7%, followed by Qwen3.7-Max (90.4%) and Gemini 3.5 Flash (77.3%).
Visual analysis
Three views of the same sourced leaderboard: model placement, the score curve, and descriptive provider averages. These visuals never alter ranking calculations.
7 ranked models · higher is better
| Model | Rank | Provider | Score |
|---|---|---|---|
| Qwen3.7-Plus | #1 | Alibaba Cloud | 91.7% |
| Qwen3.7-Max | #2 | Alibaba Cloud | 90.4% |
| Gemini 3.5 Flash | #3 | 77.3% | |
| Claude Opus 4.5 | #4 | Anthropic | 64.4% |
| Qwen3.6 Plus | #5 | Alibaba Cloud | 62% |
| Kimi K2.5 | #6 | Moonshot AI | 61% |
| GLM-5 | #7 | Z.AI | 60.8% |
Shows how quickly scores separate down the field
| Model | Rank | Provider | Score |
|---|---|---|---|
| Qwen3.7-Plus | #1 | Alibaba Cloud | 91.7% |
| Qwen3.7-Max | #2 | Alibaba Cloud | 90.4% |
| Gemini 3.5 Flash | #3 | 77.3% | |
| Claude Opus 4.5 | #4 | Anthropic | 64.4% |
| Qwen3.6 Plus | #5 | Alibaba Cloud | 62% |
| Kimi K2.5 | #6 | Moonshot AI | 61% |
| GLM-5 | #7 | Z.AI | 60.8% |
Average within the visible cohort · descriptive only
| Provider | Models | Average score |
|---|---|---|
| Alibaba Cloud | 3 | 81.4% |
| 1 | 77.3% | |
| Anthropic | 1 | 64.4% |
| Moonshot AI | 1 | 61% |
| Z.AI | 1 | 60.8% |
One best score per model · higher is better
| Rank | Model | Provider | License | Score |
|---|---|---|---|---|
| #1 | Qwen3.7-Plus qwen3.7-plus | Alibaba Cloud | — | 91.7% |
| #2 | Qwen3.7-Max qwen3.7-max | Alibaba Cloud | — | 90.4% |
| #3 | Gemini 3.5 Flash gemini-3.5-flash | closed | 77.3% | |
| #4 | Claude Opus 4.5 claude-opus-4-5 | Anthropic | closed | 64.4% |
| #5 | Qwen3.6 Plus qwen3-6-plus | Alibaba Cloud | closed | 62% |
| #6 | Kimi K2.5 kimi-k2.5 | Moonshot AI | — | 61% |
| #7 | GLM-5 glm-5 | Z.AI | open | 60.8% |
The top of this snapshot is led by Qwen3.7-Plus at 91.7%; third place is 14.4 points behind. The top-7 spread is 30.9 points.
About LongBench v2
A long-context reasoning benchmark built around demanding real-world tasks. 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
A long-context reasoning benchmark built around demanding real-world tasks.
Qwen3.7-Plus by Alibaba Cloud currently leads with 91.7%.
7 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.
