Z.ai GLM 5.1
Z.aiFlagship
ThinkingTool UseStructured Output
About this model
GLM-5.1 delivers a major leap in coding capability, with particularly significant gains in handling long-horizon tasks. Unlike previous models built around minute-level interactions, GLM-5.1 can work independently and continuously on...
Performance Tier
Flagship
Z.ai GLM 5.1 is a flagship model from Z.ai : the most capable in their lineup.
Best-in-class model from this provider. Highest performance across benchmarks, ideal for demanding tasks.
Pricing
This model is included in Elosia plansModerate
Moderate cost. A balanced choice for regular use without constant cap watching.
| Type | per 1M tokens |
|---|---|
| Input (prompt) | $1.05 |
| Output (completion) | $3.50 |
| Cache read | $0.525 |
Capabilities
Context Length203K
Max Output Tokens66K
TokenizerOther
Inputtext
Outputtext
Release DateApril 7, 2026
Benchmarks
General Intelligence
MMLU
Not reported
GPQA Diamond
86.2%
Mathematics
MATH-500
Not reported
Programming
HumanEval
Not reported
SWE-bench Verified
77.8%
LiveCodeBench
52%
Reasoning
IFEval
Not reported
Humanity's Last Exam
52.3%
Agentic
SWE-bench Pro
58.4%
Terminal-Bench 2.0
69%
Recommended Use Cases
CodingAnalysisResearchGeneral Chat
Strengths
- 754B MoE model (40B active), #1 on SWE-Bench Pro, beating Claude Opus 4.6 and GPT-5.4
- Long-horizon agentic coding, sustains autonomous execution for up to 8 hours on a single task
- 200K context window with up to 131K output tokens
- Open-weight MIT license, self-hostable, trained entirely on Huawei chips
- Competitive pricing ($0.95/M input, $3.15/M output) vs Western frontier models
Limitations
- Text-only : no multimodal (image/audio) capabilities
- Smaller ecosystem of integrations and tooling outside Chinese market
- Self-reported benchmark scoring, independent verification still limited
Resources
This model may use your data for training
