GLM · 发布历史
智谱 GLM 系列已收录 7 次发布和 143 个测试版本。
GLM 是智谱的模型品牌;Z.AI 是相关国际产品与文档品牌。完整厂商名称为北京智谱华章科技股份有限公司(Zhipu AI);官方开放平台 HTML 元数据作者及 zai-org GitHub/Hugging Face 组织与此归属相符。服务条款另列新加坡运营实体,不据此替换模型厂商归属。
全部版本
已收录测试 · 143
- Terminal-Bench · GLM-5.3模型卡;3.0GLM-5.3
- DeepSWE (v1.1) · GLM-5.3模型卡;v1.1GLM-5.3
- Agents' Last Exam (ALE-CLI) · GLM-5.3模型卡;ALE-CLIGLM-5.3
- AutomationBench (v1.0.6) · GLM-5.3模型卡;v1.0.6GLM-5.3
- HLE w/ Tools · GLM-5.3模型卡;w/ Tools; full/text-only subset not statedGLM-5.3
- GDPval-AA v2 · GLM-5.3模型卡;v2GLM-5.3
- Terminal-Bench · 2.1;GLM 官方图仅列数值,具体 score / mean reward 指标未注明GLM-5.3、GLM-5.2、GLM-5.1
- NL2Repo · GLM-5.3模型卡;unknownGLM-5.3
- ProgramBench (Almost Solved) · GLM-5.3模型卡;Almost SolvedGLM-5.3
- FrontierSWE · Release/version unspecified in GLM-5.3 cardGLM-5.3
- SWE-Marathon (v1.1) · GLM-5.3模型卡;v1.1GLM-5.3
- PostTrainBench · GLM-5.3模型卡;unknownGLM-5.3
- Toolathlon-Verified · June 2026 finalized releaseGLM-5.3
- CyberGym · GLM-5.3模型卡;unknownGLM-5.3
- ExploitGym (2h / 6h) · GLM-5.3模型卡;2h / 6hGLM-5.3
- ExploitBench · GLM-5.3模型卡;unknownGLM-5.3
- SWE-bench Pro · GLM-5.2模型卡;ProGLM-5.2
- HLE · GLM-5.2模型卡;text-only; full set not indicatedGLM-5.2
- HLE (w/ Tools) · GLM-5.2模型卡;text-only; w/ Tools; full set not indicatedGLM-5.2
- CritPt · GLM-5.2模型卡;unknownGLM-5.2
- AIME 2026 · GLM-5.2模型卡;2026; full I/II subset not statedGLM-5.2
- HMMT Nov. 2025 · GLM-5.2模型卡;Nov. 2025GLM-5.2
- HMMT Feb. 2026 · GLM-5.2模型卡;Feb. 2026GLM-5.2
- IMOAnswerBench · GLM-5.2模型卡;unknownGLM-5.2
- GPQA-Diamond · GLM-5.2模型卡;DiamondGLM-5.2
- NL2Repo · GLM-5.2模型卡;unknownGLM-5.2
- DeepSWE · GLM-5.2模型卡;unknownGLM-5.2
- ProgramBench · GLM-5.2模型卡;unknownGLM-5.2
- Terminal-Bench · GLM-5.2模型卡;2.1GLM-5.2
- FrontierSWE (Dominance) · GLM-5.2模型卡;DominanceGLM-5.2
- PostTrainBench · GLM-5.2模型卡;unknownGLM-5.2
- SWE-Marathon · GLM-5.2模型卡;unknownGLM-5.2
- MCP-Atlas (Public Set) · GLM-5.2模型卡;Public SetGLM-5.2
- Tool-Decathlon · GLM-5.2模型卡;unknownGLM-5.2
- SWE-Bench Pro · GLM-5.1模型卡;ProGLM-5.1
- HLE · GLM-5.1模型卡;unknown; full/text-only subset not statedGLM-5.1
- HLE (w/ Tools) · GLM-5.1模型卡;w/ Tools; full/text-only subset not statedGLM-5.1
- AIME 2026 · GLM-5.1模型卡;2026; full I/II subset not statedGLM-5.1
- HMMT Nov. 2025 · GLM-5.1模型卡;Nov. 2025GLM-5.1
- HMMT Feb. 2026 · GLM-5.1模型卡;Feb. 2026GLM-5.1
- IMOAnswerBench · GLM-5.1模型卡;unknownGLM-5.1
- GPQA-Diamond · GLM-5.1模型卡;DiamondGLM-5.1
- NL2Repo · GLM-5.1模型卡;unknownGLM-5.1
- Terminal-Bench · GLM-5.1模型卡;2.0GLM-5.1
- CyberGym · GLM-5.1模型卡;unknownGLM-5.1
- BrowseComp · GLM-5.1模型卡;unknownGLM-5.1
- BrowseComp (w/ Context Manage) · GLM-5.1模型卡;w/ Context ManageGLM-5.1
- τ³-Bench · GLM-5.1模型卡;unknownGLM-5.1
- MCP-Atlas (Public Set) · GLM-5.1模型卡;Public SetGLM-5.1
- Tool-Decathlon · GLM-5.1模型卡;unknownGLM-5.1
- Vending Bench 2 · GLM-5.1模型卡;2GLM-5.1
- SWE-bench Verified · GLM-5固定v2报告/模型卡;题集修订未注明GLM-5
- Terminal-Bench · 2.0; GLM source does not name score/mean-reward metricGLM-5、GLM-4.6
- Vending-Bench · 2GLM-5
- Humanity's Last Exam · GLM-5固定v2报告/模型卡;text-only subsetGLM-5
- Humanity's Last Exam · GLM-5固定v2报告/模型卡;text-only · w/ToolsGLM-5
- AIME · GLM-5固定v2报告/模型卡;2026 IGLM-5
- HMMT · GLM-5固定v2报告/模型卡;Feb. 2025GLM-5
- HMMT · GLM-5固定v2报告/模型卡;Nov. 2025GLM-5
- IMO-AnswerBench · GLM-5固定v2报告/模型卡;版本未注明GLM-5
- GPQA · GLM-5固定v2报告/模型卡;DiamondGLM-5
- LongBench · GLM-5固定v2报告/模型卡;v2GLM-5
- SWE-bench Multilingual · GLM-5固定v2报告/模型卡;题集修订未注明GLM-5
- Terminal-Bench · GLM-5固定v2报告/模型卡;2.0 · 原题集GLM-5
- Terminal-Bench · GLM-5固定v2报告/模型卡;2.0 · 修订题集GLM-5
- CyberGym · GLM-5固定v2报告/模型卡;版本未注明GLM-5
- BrowseComp · GLM-5固定v2报告/模型卡;without context managementGLM-5
- BrowseComp-ZH · GLM-5固定v2报告/模型卡;版本未注明GLM-5
- τ²-Bench · GLM-5固定v2报告/模型卡;版本未注明GLM-5
- MCP-Atlas · GLM-5固定v2报告/模型卡;500-task public setGLM-5
- GDPval-AA · GLM-5固定v2报告/模型卡;snapshot recorded 2026-02-15GLM-5
- SWE-rebench · GLM-5固定v2报告/模型卡;January 2026GLM-5
- LMArena Text · GLM-5固定v2报告/模型卡;报告快照日期未注明GLM-5
- LMArena Code · GLM-5固定v2报告/模型卡;报告快照日期未注明GLM-5
- BrowseComp · GLM-5固定v2报告/模型卡;with context managementGLM-5
- Tool-Decathlon · GLM-5固定v2报告/模型卡;版本未注明GLM-5
- CC-Bench-V2 · GLM-5固定v2报告/模型卡;v2 · HTMLGLM-5
- CC-Bench-V2 · GLM-5固定v2报告/模型卡;v2 · ReactGLM-5
- CC-Bench-V2 · GLM-5固定v2报告/模型卡;v2 · VueGLM-5
- CC-Bench-V2 · GLM-5固定v2报告/模型卡;v2 · SvelteGLM-5
- CC-Bench-V2 · GLM-5固定v2报告/模型卡;v2 · Next.jsGLM-5
- CC-Bench-V2 · GLM-5固定v2报告/模型卡;v2 · EngineeringGLM-5
- CC-Bench-V2 · GLM-5固定v2报告/模型卡;v2 · Repo ExplorationGLM-5
- CC-Bench-V2 · GLM-5固定v2报告/模型卡;v2 · Chained TasksGLM-5
- CC-Bench-V2 · GLM-5固定v2报告/模型卡;v2 · 整体GLM-5
- ZMultiTransBench · GLM-5固定v2报告/模型卡;版本未注明GLM-5
- MENT-SNS · GLM-5固定v2报告/模型卡;版本未注明GLM-5
- Human Scoring · GLM-5固定v2报告/模型卡;版本未注明GLM-5
- ZMultiDialBench · GLM-5固定v2报告/模型卡;版本未注明GLM-5
- IF-Badcase · GLM-5固定v2报告/模型卡;版本未注明GLM-5
- IFBench · GLM-5固定v2报告/模型卡;版本未注明GLM-5
- Multi Challenge · GLM-5固定v2报告/模型卡;版本未注明GLM-5
- SimpleQA · GLM-5固定v2报告/模型卡;版本未注明GLM-5
- Chinese SimpleQA · GLM-5固定v2报告/模型卡;版本未注明GLM-5
- Toolcall-Badcase · GLM-5固定v2报告/模型卡;版本未注明GLM-5
- Slide generation human evaluation · GLM-5固定v2报告/模型卡;版本未注明GLM-5
- SWE-bench Verified · GLM-4.7模型卡;VerifiedGLM-4.7
- SWE-bench Multilingual · GLM-4.7模型卡;MultilingualGLM-4.7
- LiveCodeBench · v6; GLM source metric unit unspecifiedGLM-4.7、GLM-4.6
- Terminal-Bench · GLM-4.7模型卡;2.0GLM-4.7
- Humanity's Last Exam · Release/task set unspecified; percentage explicitly stated in GLM-4.7 proseGLM-4.7
- BrowseComp · Protocol/version unspecified in GLM sourceGLM-4.7、GLM-4.6
- τ²-Bench · Domain/version/metric unspecified in GLM sourceGLM-4.7
- MMLU-Pro · GLM-4.7模型卡;unknownGLM-4.7
- GPQA-Diamond · GLM-4.7模型卡;DiamondGLM-4.7
- HLE · GLM-4.7模型卡;unknown; full/text-only subset not statedGLM-4.7
- HLE (w/ Tools) · GLM-4.7模型卡;w/ Tools; full/text-only subset not statedGLM-4.7
- AIME 2025 · GLM-4.7模型卡;2025; full I/II subset not statedGLM-4.7
- HMMT Feb. 2025 · GLM-4.7模型卡;Feb. 2025GLM-4.7
- HMMT Nov. 2025 · GLM-4.7模型卡;Nov. 2025GLM-4.7
- IMOAnswerBench · GLM-4.7模型卡;unknownGLM-4.7
- Terminal Bench Hard · GLM-4.7模型卡;HardGLM-4.7
- BrowseComp · GLM-4.7模型卡;unknownGLM-4.7
- BrowseComp (w/ Context Manage) · GLM-4.7模型卡;w/ Context ManageGLM-4.7
- BrowseComp-Zh · GLM-4.7模型卡;ZhGLM-4.7
- τ²-Bench · GLM-4.7模型卡;unknownGLM-4.7
- AIME 25 · GLM-4.6模型卡;2025; full I/II subset not statedGLM-4.6
- AIME 25 · GLM-4.6模型卡;2025; full I/II subset not stated; w/ ToolsGLM-4.6
- GPQA · Subset/version unspecified in GLM sourceGLM-4.6
- Humanity's Last Exam · Version, tools and unit unspecified in chartGLM-4.6
- SWE-bench Verified · GLM-4.6模型卡;VerifiedGLM-4.6
- τ²-Bench · Weighted aggregate; release/version unspecifiedGLM-4.6
- CC-Bench · V1.1; 74 real-world coding tests in Claude Code; exact task release unspecifiedGLM-4.6
- TAU-Bench · Retail category; release unspecifiedGLM-4.5
- TAU-Bench · Airline category; release unspecifiedGLM-4.5
- Berkeley Function Calling Leaderboard · v3GLM-4.5
- BrowseComp · GLM-4.5 report; protocol/version unspecifiedGLM-4.5
- AIME · 2024 editionGLM-4.5
- GPQA · Official GLM-4.5 report; subset/version not statedGLM-4.5
- Humanity's Last Exam · Text questions; exact release unspecifiedGLM-4.5
- LiveCodeBench · Problems from 2024-07-01 through 2025-01-01GLM-4.5
- SWE-bench Verified · Verified subset (500 problems)GLM-4.5
- Terminal-Bench · Version unspecified; official report states a percentageGLM-4.5
- MMLU · GLM-4.5 Technical Report Table 6;精确题集修订未注明,单位未注明;指标按§4.2.4原文。GLM-4.5
- SimpleQA · GLM-4.5 Technical Report Table 6;精确题集修订未注明,单位未注明;指标按§4.2.4原文。GLM-4.5
- IFEval · GLM-4.5 Technical Report Table 6;精确题集修订未注明,单位未注明;指标按§4.2.4原文。GLM-4.5
- SysBench · GLM-4.5 Technical Report Table 6;精确题集修订未注明,单位未注明;指标按§4.2.4原文。GLM-4.5
- MultiChallenge · GLM-4.5 Technical Report Table 6;精确题集修订未注明,单位未注明;指标按§4.2.4原文。GLM-4.5
- CC-Bench · 52 programming/development tasks across six domains; task release unspecifiedGLM-4.5
- LLM Performance Evaluation composite · GLM-4.5模型卡;12 benchmarks: MMLU-Pro, AIME 24, MATH-500, SciCode, GPQA, HLE, LCB (2407-2501), SWE-Bench Verified, Terminal-Bench, TAU-Bench, BFCL V3, BrowseCompGLM-4.5
- Agentic · GLM-4.5模型卡;Agentic Benchmarks: TAU-Bench, BFCL V3 (Full), BrowseCompGLM-4.5
- Reasoning · GLM-4.5模型卡;Reasoning Benchmarks: MMLU-Pro, AIME 24, MATH 500, SciCode, GPQA, HLE, LCB (2407-2501)GLM-4.5
- Coding · GLM-4.5模型卡;Coding Benchmarks: SWE-Bench Verified, Terminal-BenchGLM-4.5