GLM-5vsGLM-4.7
Across 15 shared benchmarks, GLM-5 leads overall: GLM-5 wins 12, GLM-4.7 wins 2, with 1 ties and an average score difference of +17.25.
GLM-512 wins(80%)Ties1(13%)2 winsGLM-4.7
Benchmark scores
Grouped by capability, sorted by largest gap within each. 15 shared benchmarks.
Agent Level Benchmark
GLM-5 3/4| Benchmark | GLM-5 | GLM-4.7 | Diff |
|---|---|---|---|
| Terminal Bench Hard | 39.4053 / 244Normal (With Tools) | 30.30110 / 244Normal (With Tools) | +9.10 |
| τ²-Bench - Telecom | 97.4016 / 264Normal (With Tools) | 94.2038 / 264Normal (With Tools) | +3.20 |
| τ³-Banking | 9.79130 / 164Thinking (With Tools) | 12.20122 / 164Thinking (With Tools) | -2.41 |
| τ²-Bench | 89.704 / 44Thinking (With Tools) | 87.406 / 44Thinking (With Tools) | +2.30 |
General Knowledge
GLM-5 3/3| Benchmark | GLM-5 | GLM-4.7 | Diff |
|---|---|---|---|
| LiveBench | 68.8543 / 117Normal (No Tools) | 58.0980 / 117Normal (No Tools) | +10.76 |
| HLE | 7.60427 / 563Normal (No Tools) · Text only | 6.40453 / 563Normal (No Tools) · Text only | +1.20 |
| CritPt | 2127 / 200Thinking (No Tools) | 1.70131 / 200Thinking (No Tools) | +0.30 |
Math and Reasoning
GLM-4.7 1/2| Benchmark | GLM-5 | GLM-4.7 | Diff |
|---|---|---|---|
| AIME 2026 | 92.7018 / 29Thinking (No Tools) | 92.9017 / 29Thinking (No Tools) | -0.20 |
| FrontierMath - Tier 4 | 2.1056 / 80Normal (No Tools) | 2.1056 / 80Normal (No Tools) | — |
AI Agent - Information Search
GLM-5 1/1| Benchmark | GLM-5 | GLM-4.7 | Diff |
|---|---|---|---|
| BrowseComp | 75.9026 / 57Thinking (With Tools) | 5243 / 57Thinking (With Tools) | +23.90 |
AI Agent - Tool Usage
GLM-5 1/1| Benchmark | GLM-5 | GLM-4.7 | Diff |
|---|---|---|---|
| Terminal Bench 2.0 | 61.1018 / 48Thinking (With Tools) | 4145 / 48Thinking (With Tools) | +20.10 |
General Evaluation
GLM-5 1/1| Benchmark | GLM-5 | GLM-4.7 | Diff |
|---|---|---|---|
| GPQA Diamond | 66.60349 / 462Normal (No Tools) | 66.40350 / 462Normal (No Tools) | +0.20 |
Instruction Following
GLM-5 1/1| Benchmark | GLM-5 | GLM-4.7 | Diff |
|---|---|---|---|
| IF Bench | 55.20132 / 282Normal (No Tools) | 54.60138 / 282Normal (No Tools) | +0.60 |
Long Context
GLM-5 1/1| Benchmark | GLM-5 | GLM-4.7 | Diff |
|---|---|---|---|
| AA-LCR | 43.70151 / 170Normal (No Tools) | 40.70157 / 170Normal (No Tools) | +3 |
Writing and Creative Capabilities
GLM-5 1/1| Benchmark | GLM-5 | GLM-4.7 | Diff |
|---|---|---|---|
| Creative Writing | 1,59839 / 106Normal (No Tools) | 1,41161 / 106Normal (No Tools) | +186.70 |
Specs
| Field | GLM-5 | GLM-4.7 |
|---|---|---|
| Publisher | 智谱AI | 智谱AI |
| Release date | 2026-02-11 | 2025-12-22 |
| Model type | Chat model | Chat model |
| Architecture | MoE | MoE |
| Parameters | 744B | 358B |
| Context length | 200K | 200K |
| Max output | 128K | 132072 |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | GLM-5 | GLM-4.7 |
|---|---|---|
| Text input | $1 / 1M tokens | ¥4 / 1M tokens |
| Text output | $3.2 / 1M tokens | ¥16 / 1M tokens |
| Cache read | Not public | ¥2 / 1M tokens |
| Cache write | $0.2 / 1M tokens | ¥0 / 1M tokens |
Summary
- GLM-5leads in:Agent Level Benchmark (3/4), General Knowledge (3/3), AI Agent - Information Search (1/1), AI Agent - Tool Usage (1/1), General Evaluation (1/1), Instruction Following (1/1), Long Context (1/1), Writing and Creative Capabilities (1/1)
- GLM-4.7leads in:Math and Reasoning (1/2)
On average across the 15 shared benchmarks, GLM-5 scores 17.25 higher.
Largest single-benchmark gap: Creative Writing — GLM-5 1,598 vs GLM-4.7 1,411 (+186.70).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.