Qwen3.8 Max 0902
2026-09-02 upgraded snapshot of Qwen3.8 Max with stronger coding, collaborative agents, and multimodal document understanding
Shogo Agent Index
UnrankedNot ranked yet: 1 agentic benchmark measured, and we need at least 2 plus an overall capability score. It will be ranked as results are published.
Agentic coding
55
14% measured
Long-horizon autonomy
59
estimated
Knowledge-work agents
40
estimated
Computer use
38
estimated
Security & context learning
50
estimated
Agentic benchmarks
| Benchmark | Score | Normalised | Index weight | Harness / source |
|---|---|---|---|---|
| SWE-bench Verified | est. | 8% | predicted 79 | |
| Terminal-Bench 2.0 | est. | 8% | predicted 76 | |
| DeepSWE | est. | 6% | predicted 52 | |
| FrontierSWE | 17.8% | 5% | proximus · $51.27/task | |
| CursorBench | est. | 4% | predicted 39 | |
| GSO-Bench | est. | 4% | predicted 33 | |
| METR Time Horizons | est. | 10% | predicted 76 | |
| Vending-Bench 2 | est. | 6% | predicted 53 | |
| PostTrainBench | est. | 4% | predicted 26 | |
| APEX-Agents | est. | 8% | predicted 48 | |
| GDPval | est. | 7% | predicted 49 | |
| Remote Labor Index | est. | 5% | predicted 5 | |
| DeepResearch Bench | est. | 5% | predicted 48 | |
| OSWorld 2.0 | est. | 6% | predicted 11 | |
| OSWorld | est. | 4% | predicted 79 | |
| Cybench | est. | 5% | predicted 77 | |
| CL-bench | est. | 5% | predicted 22 |
Other benchmarks
| Benchmark | Category | Score | Source |
|---|---|---|---|
| Chess Puzzles | Reasoning | 37.0% | source |
| CritPt | Reasoning | 17.7% | source |
| FrontierMath Tiers 1-3 | Math | 65.6% | source |
| FrontierMath Tier 4 | Math | 34.1% | source |
| GPQA Diamond | Reasoning | 92.3% | source |
| MirrorCode | Coding | 47.8% | source |
| OTIS Mock AIME | Math | 100.0% | source |
| SciCode | Coding | 52.1% | source |
| SimpleQA Verified | Knowledge | 47.3% | source |