Beijing Zhipu Huazhang Technology Co., Ltd. recently announced that it will provide the API interface for its GLM-4-Flash large language model free to the public to promote the popularization and application of large model technology. The GLM-4-Flash model shows significant advantages in speed and performance, particularly in inference speed, by employing adaptive weight quantization and parallel processing technologies.