Model
GLM-5.2
Open flagship GLM for long-horizon coding agents and million-token context work
1 providersreleased open weights
$1.40
best input /M · Zhipu AI
$4.40
best output /M
1M
context window
131K
max output
Compare providers (sorted by listed input price)
| Provider | Input /M | Output /M | Cache read /M | Cache write /M | Context | Capabilities | Status | Updated |
|---|---|---|---|---|---|---|---|---|
| Zhipu AI glm-5.2 | $1.40 | $4.40 | $0.26 | $0 | 1M | reasoningtoolsstructuredno vision | Jun 13, 2026 |
Price badge
https://models.sutraworks.ai/badge/zhipuai/glm-5.2.svgLive SVG, regenerated on every hourly sync — paste it into a README to always show the current cheapest listed price. Updates when prices change; no build step needed on your side.
Benchmarks
- SWE-Bench Pro ↗(resolve rate)report ↗62.1
- Terminal-Bench ↗(success rate)report ↗82.7
- FrontierSWE ↗(dominance)report ↗74.4
- Humanity's Last Exam ↗(accuracy)report ↗40.5
- Humanity's Last Exam ↗(accuracy)report ↗54.7
- CritPt(accuracy)report ↗20.9
- AIME(accuracy)report ↗99.2
- HMMT(accuracy)report ↗94.4
- HMMT(accuracy)report ↗92.5
- IMOAnswerBench(accuracy)report ↗91
- GPQA Diamond ↗(accuracy)report ↗91.2
- NL2Repo(resolve rate)report ↗48.9
- DeepSWE(resolve rate)report ↗46.2
- Program Bench(score)report ↗63.7
- Terminal-Bench ↗(success rate)report ↗81
- PostTrainBench(score)report ↗34.3
- SWE Marathon ↗(resolve rate)report ↗13
- MCP Atlas(score)report ↗76.8
- Tool-Decathlon(score)report ↗48.2
Weights
Prices are per million tokens (USD). “—” means the provider does not publicly list a price for this model. Data from models.dev; verify with the provider before purchasing.