Zhipu AI Unveils GLM-4.6: Bigger Context Window, Improved Efficiency
Zhipu AI has launched GLM-4.6, the latest iteration of its GLM series, boasting significant improvements and a larger context window. The model aims to reduce token consumption in practical tasks and is licensed under the MIT license.
GLM-4.6 offers a substantial increase in its input window, now supporting 200K tokens, double that of its predecessor GLM-4.5. This expansion allows the model to excel in tasks requiring extended input, such as coding and long-form text generation.
On the extended CC-Bench, GLM-4.6 performs near parity with Claude Sonnet 4, a model renowned for its high performance. Notably, GLM-4.6 uses 15% fewer tokens than GLM-4.5, indicating improved efficiency. The model has 355B parameters and supports BF16/F32 tensors.
GLM-4.6 is available via Z.ai API and OpenRouter, making it accessible for integration with popular coding agents. It also supports local serving with vLLM and SGLang, providing flexibility for users. Weights for the model are openly available on Hugging Face and ModelScope, enabling local deployment.
Zhipu AI's GLM-4.6 is a significant update to the GLM series, offering a larger context window and improved performance. With open weights and integration capabilities, it is poised to make a substantial impact in tasks requiring extended input and efficient token consumption.
Read also:
- Hematology specialist and anemia treatment: The role of a hematologist in managing anemia conditions
- Trump announces Chinese leader's confirmation of TikTok agreement
- Hackers Utilize GOLD SALEM to Infiltrate Networks and Evade Security Measures, Deploying Warlock Ransomware
- Southeast Asia's initial AI-linked battery-swapping station was launched by U Power
