Qwen3.8-Max-0902vsGLM-5.3
Across 7 shared benchmarks, Qwen3.8-Max-0902 leads overall: Qwen3.8-Max-0902 wins 7, GLM-5.3 wins 0, with 0 ties and an average score difference of +3.46.
Qwen3.8-Max-0902
阿里巴巴 · 2026-09-02 · Reasoning model
GLM-5.3
智谱AI · 2026-08-14 · Reasoning model
Qwen3.8-Max-09027 wins(100%)(0%)0 winsGLM-5.3
Benchmark scores
Grouped by capability, sorted by largest gap within each. 7 shared benchmarks.
Coding and Software Engineer
Qwen3.8-Max-0902 4/4| Benchmark | Qwen3.8-Max-0902 | GLM-5.3 | Diff |
|---|---|---|---|
| Program Bench | 285 / 7极高强度思考（工具） | 196 / 7Max (With Tools) | +9 |
| NL2Repo-Bench | 64.901 / 12极高强度思考（工具） | 584 / 12Max (With Tools) | +6.90 |
| DeepSWE | 69.304 / 32极高强度思考（工具） | 66.909 / 32Max (With Tools) | +2.40 |
| SWE-Marathon | 44.801 / 6极高强度思考（工具） | 42.502 / 6Max (With Tools) | +2.30 |
AI Agent - Tool Usage
Qwen3.8-Max-0902 3/3| Benchmark | Qwen3.8-Max-0902 | GLM-5.3 | Diff |
|---|---|---|---|
| AutomationBench | 50.801 / 12极高强度思考（工具） | 48.203 / 12Max (With Tools) | +2.60 |
| Terminal-Bench 3.0 | 293 / 7极高强度思考（工具） | 28.304 / 7Max (With Tools) | +0.70 |
| Toolathlon-Verified | 73.307 / 10极高强度思考（工具） | 738 / 10Max (With Tools) | +0.30 |
Specs
| Field | Qwen3.8-Max-0902 | GLM-5.3 |
|---|---|---|
| Publisher | 阿里巴巴 | 智谱AI |
| Release date | 2026-09-02 | 2026-08-14 |
| Model type | Reasoning model | Reasoning model |
| Architecture | MoE | MoE |
| Parameters | 2.4T | 744B |
| Context length | 1M | 1M |
| Max output | 131K | 128K |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | Qwen3.8-Max-0902 | GLM-5.3 |
|---|---|---|
| Text input | $2 / 1M tokens | $1.4 / 1M tokens |
| Text output | $6 / 1M tokens | $4.4 / 1M tokens |
| Cache read | $0.25 / 1M tokens | $0.26 / 1M tokens |
| Cache write | $2.5 / 1M tokens | Not public |
Summary
- Qwen3.8-Max-0902leads in:Coding and Software Engineer (4/4), AI Agent - Tool Usage (3/3)
On average across the 7 shared benchmarks, Qwen3.8-Max-0902 scores 3.46 higher.
Largest single-benchmark gap: Program Bench — Qwen3.8-Max-0902 28 vs GLM-5.3 19 (+9).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.