GLM-5.2vsMiniMax M3
Across 6 shared benchmarks, GLM-5.2 leads overall: GLM-5.2 wins 5, MiniMax M3 wins 1, with 0 ties and an average score difference of +14.95.
GLM-5.2
智谱AI · 2026-06-13 · Reasoning model
MiniMax M3
MiniMaxAI · 2026-06-01 · Multimodal model
GLM-5.25 wins(83%)(17%)1 winMiniMax M3
Benchmark scores
Grouped by capability, sorted by largest gap within each. 6 shared benchmarks.
Coding and Software Engineer
GLM-5.2 2/3| Benchmark | GLM-5.2 | MiniMax M3 | Diff |
|---|---|---|---|
| Text Arena (Coding) | 1,5935 / 35最高（无工具） | 1,52814 / 35Normal (No Tools) | +65.50 |
| SWE-Bench Pro - Public | 62.109 / 57Thinking (With Tools) | 5913 / 57Thinking (With Tools) | +3.10 |
| PostTrain Bench | 34.304 / 4Max (With Tools) | 372 / 4Thinking (With Tools) | -2.70 |
AI Agent - Tool Usage
GLM-5.2 2/2| Benchmark | GLM-5.2 | MiniMax M3 | Diff |
|---|---|---|---|
| Terminal-Bench 2.1 | 8117 / 44Thinking High (With Tools) | 6635 / 44Thinking (With Tools) | +15 |
| MCP-Atlas | 76.8013 / 38Thinking (With Tools) | 74.2022 / 38Thinking (With Tools) | +2.60 |
General Knowledge
GLM-5.2 1/1| Benchmark | GLM-5.2 | MiniMax M3 | Diff |
|---|---|---|---|
| LiveBench | 76.249 / 115Normal (No Tools) | 70.0240 / 115Deep Thinking (No Tools) | +6.22 |
Specs
| Field | GLM-5.2 | MiniMax M3 |
|---|---|---|
| Publisher | 智谱AI | MiniMaxAI |
| Release date | 2026-06-13 | 2026-06-01 |
| Model type | Reasoning model | Multimodal model |
| Architecture | MoE | MoE |
| Parameters | 753.33B | 428B |
| Context length | 1M | 1M |
| Max output | 128K | 512K |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | GLM-5.2 | MiniMax M3 |
|---|---|---|
| Text input | $1.4 / 1M tokens | ¥2.1 / 1M tokens |
| Text output | $4.4 / 1M tokens | ¥8.4 / 1M tokens |
| Cache read | $0.26 / 1M tokens | ¥0.42 / 1M tokens |
Summary
- GLM-5.2leads in:Coding and Software Engineer (2/3), AI Agent - Tool Usage (2/2), General Knowledge (1/1)
On average across the 6 shared benchmarks, GLM-5.2 scores 14.95 higher.
Largest single-benchmark gap: Text Arena (Coding) — GLM-5.2 1,593 vs MiniMax M3 1,528 (+65.50).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.