Z AI
GLM 5
GLM-5 is Z.ai’s flagship open-source foundation model engineered for complex systems design and long-horizon agent workflows. Built for expert developers, it delivers production-grade performance on large-scale programming tasks, rivaling leading closed-source models. With advanced agentic planning, deep backend reasoning, and iterative self-correction, GLM-5 moves beyond code generation to full-system construction and autonomous execution.
- Input / 1M tokens
- $0.650
- Output / 1M tokens
- $2.08
- Context window
- 203K tokens
- Provider
- Z AI
- Cached input / 1M
- $0.130
Performance
Median streaming throughput and first-token latency measured by Artificial Analysis.
- Output tokens / sec
- 63 t/s
- Time to first token
- 0.88s
Benchmarks
Intelligence, coding, and math indexes plus the underlying evaluation scores.
- Intelligence Index
- 50
- Coding Index
- 44
- Math Index
- —
- MMLU-Pro
- —
- GPQA
- 82.0%
- HLE
- 27.2%
- LiveCodeBench
- —
- SciCode
- 46.2%
- MATH-500
- —
- AIME
- —
Benchmarks via Artificial Analysis