LLMStack is a full-featured foundation that lets developers build AI applications, including chatbots and agents, using pre-trained language models from companies like OpenAI and Hugging Face. It includes a no-code builder for integrating with data and business processes, and supports multi-tenancy and permission controls. LLMStack can run in the cloud or in your own data center, so it's good for a wide range of use cases.
If you're looking for something more focused, Predibase is a relatively inexpensive foundation for fine-tuning and serving large language models. It supports a range of models, including quantization and low-rank adaptation for efficient fine-tuning. Predibase also offers free serverless inference for up to 1 million tokens per day and enterprise-grade security.
Another contender is LMSYS Org, which offers a variety of tools for training and evaluating large language models, including the open-source chatbot Vicuna and the FastChat platform for training, serving and evaluating LLM-based chatbots. The organization also offers large-scale datasets and evaluation tools, making it easier to develop and deploy next-gen chatbot technology.
If you're looking for something more extensible and enterprise-ready, Microsoft Bot Framework is an open-source foundation that taps into Azure Cognitive Services for sophisticated natural language understanding and generation. It includes a visual authoring canvas and a range of AI abilities, making it good for building virtual assistants, customer service and enterprise apps.