Inflection AI Debuts Inflection-2 LLM

Inflection AI has introduced Inflection-2, the latest iteration of its large language model (LLM), which the startup claims is the best model in its compute class and the second most capable LLM in the world. Inflection-2 will be available through the startup’s API and will power Inflection’s generative AI chatbot, Pi.

Inflection-2 Flex

Inflection-2 was trained on 5,000 Nvidia H100 GPUs and has shown superior performance compared to its predecessor and other leading models, such as Google’s PaLM 2 Large, in benchmarks. The model excels in areas such as factual knowledge, stylistic control, and reasoning, outperforming several standard generative AI benchmark tests. Despite its larger size compared to Inflection-1, the model’s efficient design and transition to H100 GPUs have led to cost reductions and increased speed.

Inflection Ascendant

Inflection is focused on creating an emotionally intelligent AI conversationalist that can replicate empathy and human emotion. Inflection-2 will augment Inflection’s chatbot, Pi, after passing safety and utility tests. The startup’s choice to compare Inflection-2 with Google’s PaLM 2 model suggests its aim to position itself as a reasonable alternative to OpenAI. Inflection has recently raised $1.3 billion in funding, allowing it to continue its projects and work towards building the largest AI cluster in the world, consisting of 22,000 Nvidia H100 Tensor Core GPUs.

Conclusion

Inflection AI’s launch of Inflection-2, its latest large language model, offers improved performance in a variety of areas and positions the startup as a potential alternative to OpenAI. With the backing of major tech players and significant funding, Inflection is on a mission to build a personal AI for everyone and aims to create the largest AI cluster in the world. The introduction of Inflection-2 marks a significant milestone in the company’s scaling journey, with plans for even larger models in the future.