LLM Reference
Together.ai

Together.ai

Blazing-fast, cost-effective AI inference solutions

About

Together AI, a dynamic entity in the realm of generative artificial intelligence (AI) and large language models (LLMs), was established in 2022 in San Francisco. This innovative company was founded by a team of esteemed AI researchers and pioneers, including Vipul Ved Prakash, Ce Zhang, Chris Ré, and Percy Liang. These founders bring an extensive background in AI model architectures and infrastructure, driving the company's mission to democratize access to generative AI technologies. The core belief at Together AI is the power of open-source AI. By promoting transparency, privacy, and flexibility, they empower users to retain ownership of their models and data—a critical consideration amidst growing concerns about data privacy and vendor lock-in. Through decentralized cloud services, Together AI facilitates organizations in leveraging generative AI capabilities seamlessly, removing the typical complexities found in large-scale AI infrastructure. One of Together AI's notable contributions to the field is the RedPajama project, which aspires to craft chat-based AI models that rival leading platforms such as OpenAI's ChatGPT. The project boasts the RedPajama dataset, comprising over 30 trillion tokens, making it a cornerstone for training LLMs. This substantial open dataset underscores Together AI's dedication to advancing open-source research and providing valuable tools for AI progress. Technological innovation is a hallmark of Together AI, as reflected in the development of FlashAttention. This technique enhances the efficiency of the attention mechanism within transformer models, leading to marked improvements in training and inference speeds. Such advancements not only benefit academic pursuits but also translate into practical solutions that boost the capabilities of AI systems operationally. The collective expertise at Together AI spans machine learning, natural language processing, and cloud computing, among other fields. The multidisciplinary team thrives on addressing complex AI research and development challenges. With a leadership team that comprises individuals with distinguished careers in academia and industry, Together AI persistently leads the charge in AI innovation. Ultimately, Together AI envisions a future where generative AI is not only accessible and transparent but also widely beneficial. By championing open-source initiatives and nurturing a collaborative environment, Together AI is reshaping the landscape of AI research and deployment. Their dedication to empowering developers and researchers places them at the forefront of the swiftly evolving field of generative AI and LLMs.

Model Families

Information

Founded2022
San Francisco, California, United States