Currently listed through these providers:
Model details
GLM-5.1
GLM-5.1 is Z.ai's next-generation flagship model, positioned as the successor to GLM-5 with a sharpened focus on agentic engineering and long-horizon tasks. Rather than chasing only first-pass quality, the design goal is sustained productivity on complex, multi-step coding work: the model is built to keep making progress across long sessions, revisit earlier reasoning, revise its strategy, and stay useful when problems are ambiguous or stretch across many iterations. This makes it a strong fit for developers who need an assistant that can plan, execute, run experiments, read results, and recover from blockers inside extended agent loops rather than short single-turn exchanges.
Eval evidence from the launch write-up highlights the practical payoff of that design. Z.ai reports a SWE-Bench Pro score of 58.4, framed as state-of-the-art on complex software engineering tasks and ahead of GLM-5, GPT-5.4, Opus 4.6, and Gemini 3.1 Pro on the supplied chart. The same release notes a wide margin over GLM-5 on NL2Repo repository generation and on Terminal-Bench 2.0, a real-world terminal task benchmark that rewards the same iterative, tool-using behavior the model is tuned for. Weights are openly available, which lets teams self-host and integrate the model into coding agents, automated refactoring pipelines, and other software-engineering workflows where long-context, tool-driven reasoning matters more than a polished single answer.
Quick Info
Powered by- Provider
- TokenGo
- Model key
- z-ai/glm-5.1
- Release date
- Apr 7, 2026
- Last updated
- Apr 7, 2026
- Input modalities
- Output modalities
- Capabilities
Cost
- Input token cost
- $1.40
- Output token cost
- $4.40
Limits
- Output tokens
- 131,072 tokens
- Context window
- 200,000 tokens