MiniMax M2.5vsGLM-5
Across 14 shared benchmarks, GLM-5 leads overall: MiniMax M2.5 wins 6, GLM-5 wins 7, with 1 ties and an average score difference of -2.29.
MiniMax M2.5
MiniMaxAI · 2026-02-12 · Reasoning model
GLM-5
智谱AI · 2026-02-11 · Chat model
MiniMax M2.56 wins(43%)Ties1(50%)7 winsGLM-5
Benchmark scores
Grouped by capability, sorted by largest gap within each. 14 shared benchmarks.
General Knowledge
GLM-5 3/5| Benchmark | MiniMax M2.5 | GLM-5 | Diff |
|---|---|---|---|
| HLE | 19.40121 / 172Thinking (No Tools) | 50.4025 / 172 | -31 |
| ARC-AGI | 63.7035 / 68Thinking (No Tools) | 44.7047 / 68Thinking (No Tools) | +19 |
| LiveBench | 60.1468 / 115Deep Thinking (No Tools) | 68.8543 / 115Normal (No Tools) | -8.71 |
| GPQA Diamond | 85.2053 / 187Thinking (No Tools) | 8648 / 187Thinking (No Tools) | -0.80 |
| ARC-AGI-2 | 4.9047 / 62Thinking (No Tools) | 4.9047 / 62Thinking (No Tools) | — |
Claw-style Agent Evaluation
MiniMax M2.5 2/2| Benchmark | MiniMax M2.5 | GLM-5 | Diff |
|---|---|---|---|
| Pinch Bench | 87.806 / 37Thinking (With Tools) | 86.4012 / 37Thinking (With Tools) | +1.40 |
| Claw Bench | 92.104 / 29Thinking (With Tools) | 91.705 / 29Thinking (With Tools) | +0.40 |
Agent Level Benchmark
GLM-5 1/1| Benchmark | MiniMax M2.5 | GLM-5 | Diff |
|---|---|---|---|
| τ²-Bench - Telecom | 97.8010 / 35 | 985 / 35 | -0.20 |
AI Agent - Information Search
MiniMax M2.5 1/1| Benchmark | MiniMax M2.5 | GLM-5 | Diff |
|---|---|---|---|
| BrowseComp | 76.3023 / 53 | 75.9024 / 53 | +0.40 |
AI Agent - Tool Usage
GLM-5 1/1| Benchmark | MiniMax M2.5 | GLM-5 | Diff |
|---|---|---|---|
| Terminal Bench 2.0 | 51.7031 / 47 | 61.1018 / 47 | -9.40 |
Coding and Software Engineer
MiniMax M2.5 1/1| Benchmark | MiniMax M2.5 | GLM-5 | Diff |
|---|---|---|---|
| SWE-bench Verified | 80.2014 / 112 | 77.8025 / 112Thinking (No Tools) | +2.40 |
Instruction Following
GLM-5 1/1| Benchmark | MiniMax M2.5 | GLM-5 | Diff |
|---|---|---|---|
| IF Bench | 7013 / 30 | 7211 / 30 | -2 |
Long Context
MiniMax M2.5 1/1| Benchmark | MiniMax M2.5 | GLM-5 | Diff |
|---|---|---|---|
| AA-LCR | 69.505 / 15Thinking (No Tools) | 6314 / 15Thinking (No Tools) | +6.50 |
Productivity Knowledge
GLM-5 1/1| Benchmark | MiniMax M2.5 | GLM-5 | Diff |
|---|---|---|---|
| GDPval-AA | 3617 / 21Thinking (No Tools) | 4614 / 21Thinking (No Tools) | -10 |
Specs
| Field | MiniMax M2.5 | GLM-5 |
|---|---|---|
| Publisher | MiniMaxAI | 智谱AI |
| Release date | 2026-02-12 | 2026-02-11 |
| Model type | Reasoning model | Chat model |
| Architecture | MoE | MoE |
| Parameters | 229B | 744B |
| Context length | 128K | 200K |
| Max output | Not available | 128K |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | MiniMax M2.5 | GLM-5 |
|---|---|---|
| Text input | $0.3 / 1M tokens | $1 / 1M tokens |
| Text output | $2.4 / 1M tokens | $3.2 / 1M tokens |
| Cache write | Not public | $0.2 / 1M tokens |
Summary
- MiniMax M2.5leads in:Claw-style Agent Evaluation (2/2), AI Agent - Information Search (1/1), Coding and Software Engineer (1/1), Long Context (1/1)
- GLM-5leads in:General Knowledge (3/5), Agent Level Benchmark (1/1), AI Agent - Tool Usage (1/1), Instruction Following (1/1), Productivity Knowledge (1/1)
On average across the 14 shared benchmarks, GLM-5 scores 2.29 higher.
Largest single-benchmark gap: HLE — MiniMax M2.5 19.40 vs GLM-5 50.40 (-31).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.