Back to glossary

Term

GLM-5.1

GLM-5.1 is Z.ai's open-source flagship from April 2026 — 744B parameters (40B active), 200k context, designed for agentic coding with up to 8 hours of autonomous runtime.

GLM-5.1 — explained in more detail

Z.ai (formerly Zhipu AI) released GLM-5.1 on April 7, 2026. It is a post-training upgrade to GLM-5 built on a Mixture-of-Experts architecture with 744 billion parameters (40B active), a 200k context window and a maximum output length of 131k tokens. The jump comes not from additional pretraining but from refined reinforcement learning and alignment — explicitly aimed at agentic engineering tasks.

Example / Practical use

On SWE-Bench Pro, GLM-5.1 scores 58.4, edging past GPT-5.4 (57.7) and Claude Opus 4.6 (57.3) at launch. The model can run a single coding task autonomously for up to 8 hours — planning, execution, tests and optimization in a loop. Weights are available on Hugging Face under the MIT license; API pricing is $1.00 per 1M input tokens and $3.20 per 1M output tokens.

Entdecke mehr

Themenuebersicht