Qwen3.6 Max Preview
Flagship Qwen model for complex reasoning, coding, and agentic workflows
Shogo Agent Index
Estimated#55 of 9240.5range 29.2–51.9 · 14% of weight measured
Agentic coding
42
23% measured
Long-horizon autonomy
56
30% measured
Knowledge-work agents
31
estimated
Computer use
27
estimated
Security & context learning
40
estimated
Price
- Input
- $2.15/M
- Output
- $12.86/M
- Cache read
- $0.20/M
- 3:1 blend
- $4.83/M
- Agent-effective
- $1.08/M
Agent-effective = 20 input tokens per output token, 85% cache hits. Why.
Agentic benchmarks
| Benchmark | Score | Normalised | Index weight | Harness / source |
|---|---|---|---|---|
| SWE-bench Verified | 76.7% | 8% | source | |
| Terminal-Bench 2.0 | est. | 8% | predicted 53 | |
| DeepSWE | est. | 6% | predicted 35 | |
| FrontierSWE | est. | 5% | predicted 8 | |
| CursorBench | est. | 4% | predicted 31 | |
| GSO-Bench | est. | 4% | predicted 16 | |
| METR Time Horizons | est. | 10% | predicted 70 | |
| Vending-Bench 2 | $4,254 | 6% | source | |
| PostTrainBench | est. | 4% | predicted 21 | |
| APEX-Agents | est. | 8% | predicted 34 | |
| GDPval | est. | 7% | predicted 38 | |
| Remote Labor Index | est. | 5% | predicted 2 | |
| DeepResearch Bench | est. | 5% | predicted 46 | |
| OSWorld 2.0 | est. | 6% | predicted 5 | |
| OSWorld | est. | 4% | predicted 61 | |
| Cybench | est. | 5% | predicted 61 | |
| CL-bench | est. | 5% | predicted 19 |
Other benchmarks
| Benchmark | Category | Score | Source |
|---|---|---|---|
| DTBench | Knowledge | 87.2% | source |
| GPQA Diamond | Reasoning | 87.4% | source |
| LMCA | Knowledge | 42.5% | source |
| OTIS Mock AIME | Math | 91.1% | source |
| SimpleBench | Reasoning | 63.0% | source |
| SimpleQA Verified | Knowledge | 52.0% | source |
| WebDev Arena | Coding | 1479 | source |