Model dossier模型档案
GLM 5.1
GLM 5.1 is a mid-tier model from Z.ai (Intelligence Score 70.2/100). Strongest in agent (86.5) and knowledge (78.5); best fit: tool-use and multi-step agents. Mid-priced: $0.966 in / $3.04 out per 1M tokens. Open weights — self-hostable. Reachable via CN-direct endpoints.
GLM 5.1 是 Z.ai 的中端梯队模型(智能评分 70.2/100)。 最强项是智能体(86.5分),其次是知识(78.5分);适合工具调用与多步 Agent。 定价属中档定价:每百万 token 输入 $0.966 / 输出 $3.04。 开放权重,可自托管。 支持国内直连。
open weights开放权重 ⌂ CN direct国内直连
70.2
Intelligence Score · benchmark coverage 智能评分 · benchmark 覆盖 100%
◈
Benchmark scoresBenchmark 成绩
11 entries 条| Benchmark基准 | Score成绩 | Index指数 | Dated日期 | Measured by测评方 | |
|---|---|---|---|---|---|
| Humanity's Last ExamReasoning | 23.68 | 41 | — | official官方榜 | source ↗×4⚠±35.42 |
| GPQA DiamondReasoning | 87⚡high | 91 | — | 3rd-party第三方 | source ↗×5⚠±3.7 |
| MMLU-ProKnowledge | 72.7 | 79 | — | 3rd-party第三方 | source ↗×2⚠±9.3 |
| SWE-bench VerifiedCoding | 58.4 | 61 | — | 3rd-party第三方 | source ↗×2 |
| AIME 2025Math | 95.3 | 95 | — | 3rd-party第三方 | source ↗×4 |
| LiveCodeBenchCoding | 70.6 | 76 | — | 3rd-party第三方 | source ↗×4⚠±18.8 |
| FrontierMathMath | 33.448 | 36 | — | 3rd-party第三方 | source ↗×2⚠±25.55 |
| Terminal-BenchCoding | 63.5 | 69 | 2026-07 | 3rd-party第三方 | source ↗×5⚠±7.3 |
| τ²-benchAgent | 97.7⚡high | 99 | 2026-06-10 | 3rd-party第三方 | source ↗×4 |
| LMArena (Chatbot Arena)Preference | 1467 | 61 | 2026-08-12 | official官方榜 | source ↗×3 |
| WebArenaAgent | 68 | 74 | — | 3rd-party第三方 | source ↗ |
≡
Specs & pricing规格与价格
| Provider厂商 | Z.ai |
|---|---|
| Released发布 | 2026-04-07 |
| Context window上下文 | 205K tokens |
| Max output最大输出 | 128K tokens |
| Modality模态 | text->text |
| Reasoning tiers推理档位 | ⚡none ⚡high default默认 high · thinking on / off |
| API input price输入价格 | $0.966 / 1M tokens |
| API output price输出价格 | $3.04 / 1M tokens |
⛓
Tools using this model使用它的工具
How we score →评分方法 → · Catalog & pricing via the official OpenRouter API.模型目录与价格来自 OpenRouter 官方 API。