Tencent Releases 770 Billion Parameter Open-Weight AI Model Under Apache 2.0 License
Chinese tech giant Tencent has released a 770 billion parameter open-weight AI model under the permissive Apache 2.0 license, making it freely available for commercial and research use. The release adds to a growing list of large open-weight models competing with proprietary systems from OpenAI and Anthropic.
Chinese technology company Tencent has released a 770 billion parameter open-weight AI model under the Apache 2.0 license, making it freely available for commercial and research use worldwide.
The release adds to a growing list of large open-weight models that are competing with proprietary systems from companies like OpenAI and Anthropic. Open-weight models allow developers to download, modify, and deploy the model without paying licensing fees or sending data to a third-party API.
The Apache 2.0 license is one of the most permissive open-source licenses available, allowing companies to use the model in commercial products without restrictions beyond attribution.
At 770 billion parameters, the model is among the largest open-weight models released to date. Larger models generally perform better on complex tasks, though they also require more computing power to run.
The release comes as the global AI industry is increasingly divided between proprietary systems controlled by a handful of large companies and open-weight models that give developers more control and flexibility.
Meta has been the most prominent US company releasing open-weight models through its Llama series. Tencent's release signals that Chinese companies are also committed to the open-weight approach, which can help build developer ecosystems and establish technical standards.
The model is expected to be particularly useful for developers in Asia who want to build AI applications in Chinese and other Asian languages, where Tencent's training data may give it an advantage over Western models.
Researchers and developers can access the model through standard AI model repositories. Tencent has not disclosed the full details of the training data or the compute used to build the model.
