Qwen3.8-Max-0902vsDeepSeek-V4-Pro
Across 4 shared benchmarks, Qwen3.8-Max-0902 leads overall: Qwen3.8-Max-0902 wins 3, DeepSeek-V4-Pro wins 1, with 0 ties and an average score difference of +7.05.
Qwen3.8-Max-0902
阿里巴巴 · 2026-09-02 · Reasoning model
DeepSeek-V4-Pro
DeepSeek-AI · 2026-08-13 · Reasoning model
Qwen3.8-Max-09023 wins(75%)(25%)1 winDeepSeek-V4-Pro
Benchmark scores
Grouped by capability, sorted by largest gap within each. 4 shared benchmarks.
AI Agent - Tool Usage
Even 2/2| Benchmark | Qwen3.8-Max-0902 | DeepSeek-V4-Pro | Diff |
|---|---|---|---|
| AutomationBench | 50.801 / 12极高强度思考（工具） | 31.805 / 12极高强度思考（工具） | +19 |
| Toolathlon-Verified | 73.307 / 10极高强度思考（工具） | 74.104 / 10极高强度思考（工具） | -0.80 |
Coding and Software Engineer
Qwen3.8-Max-0902 2/2| Benchmark | Qwen3.8-Max-0902 | DeepSeek-V4-Pro | Diff |
|---|---|---|---|
| DeepSWE | 69.304 / 32极高强度思考（工具） | 62.7014 / 32极高强度思考（工具） | +6.60 |
| NL2Repo-Bench | 64.901 / 12极高强度思考（工具） | 61.502 / 12极高强度思考（工具） | +3.40 |
Specs
| Field | Qwen3.8-Max-0902 | DeepSeek-V4-Pro |
|---|---|---|
| Publisher | 阿里巴巴 | DeepSeek-AI |
| Release date | 2026-09-02 | 2026-08-13 |
| Model type | Reasoning model | Reasoning model |
| Architecture | MoE | MoE |
| Parameters | 2.4T | 1.6T |
| Context length | 1M | 1M |
| Max output | 131K | 384K |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | Qwen3.8-Max-0902 | DeepSeek-V4-Pro |
|---|---|---|
| Text input | $2 / 1M tokens | $0.435 / 1M tokens |
| Text output | $6 / 1M tokens | $0.87 / 1M tokens |
| Cache read | $0.25 / 1M tokens | $0.003625 / 1M tokens |
| Cache write | $2.5 / 1M tokens | Not public |
Summary
- Qwen3.8-Max-0902leads in:Coding and Software Engineer (2/2)
- Tied in:AI Agent - Tool Usage
On average across the 4 shared benchmarks, Qwen3.8-Max-0902 scores 7.05 higher.
Largest single-benchmark gap: AutomationBench — Qwen3.8-Max-0902 50.80 vs DeepSeek-V4-Pro 31.80 (+19).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.