GLM-5.2vsClaude Opus 4.8
Across 4 shared benchmarks, Claude Opus 4.8 leads overall: GLM-5.2 wins 1, Claude Opus 4.8 wins 3, with 0 ties and an average score difference of -2.65.
GLM-5.2
智谱AI · 2026-06-13 · Reasoning model
Claude Opus 4.8
Anthropic · 2026-05-28 · Reasoning model
GLM-5.21 win(25%)(75%)3 winsClaude Opus 4.8
Benchmark scores
Grouped by capability, sorted by largest gap within each. 4 shared benchmarks.
General Knowledge
Claude Opus 4.8 2/2| Benchmark | GLM-5.2 | Claude Opus 4.8 | Diff |
|---|---|---|---|
| HLE | 54.708 / 159Thinking (With Tools) | 57.905 / 159Extended (with tools) | -3.20 |
| GPQA Diamond | 91.2015 / 179Thinking (No Tools) | 93.606 / 179Thinking High (No Tools) | -2.40 |
AI Agent - Tool Usage
GLM-5.2 1/1| Benchmark | GLM-5.2 | Claude Opus 4.8 | Diff |
|---|---|---|---|
| TerminalBench 2.1 | 814 / 14Thinking High (With Tools) | 78.905 / 14Thinking High (With Tools) | +2.10 |
Coding and Software Engineer
Claude Opus 4.8 1/1| Benchmark | GLM-5.2 | Claude Opus 4.8 | Diff |
|---|---|---|---|
| SWE-Bench Pro - Public | 62.105 / 44Thinking (With Tools) | 69.203 / 44Extended (with tools) | -7.10 |
Specs
| Field | GLM-5.2 | Claude Opus 4.8 |
|---|---|---|
| Publisher | 智谱AI | Anthropic |
| Release date | 2026-06-13 | 2026-05-28 |
| Model type | Reasoning model | Reasoning model |
| Architecture | MoE | Dense |
| Parameters | 753.33B | Not available |
| Context length | 1M | 1M |
| Max output | 128K | 125K |
API pricing
Prices use DataLearner records when available; missing fields are not inferred.
| Item | GLM-5.2 | Claude Opus 4.8 |
|---|---|---|
| Text input | $1.4 / 1M tokens | $5 / 1M tokens |
| Text output | $4.4 / 1M tokens | $25 / 1M tokens |
| Cache read | $0.26 / 1M tokens | $0.5 / 1M tokens |
| Cache write | Not public | $6.25 / 1M tokens |
Summary
- GLM-5.2leads in:AI Agent - Tool Usage (1/1)
- Claude Opus 4.8leads in:General Knowledge (2/2), Coding and Software Engineer (1/1)
On average across the 4 shared benchmarks, Claude Opus 4.8 scores 2.65 higher.
Largest single-benchmark gap: SWE-Bench Pro - Public — GLM-5.2 62.10 vs Claude Opus 4.8 69.20 (-7.10).
Page generated from structured model, pricing and benchmark records. No real-time LLM is used to write the prose.