ChinaIncluding Funding rounds, Bull / Bear thesis, Stock + earnings, Roster changes, Patents, News, and Open roles.
Already subscribed? Sign in →
Jiang Daxin is the founder and CEO of StepFun, the Shanghai multimodal foundation-model company he started in April 2023. He was previously an executive at Microsoft, where he worked on products including Bing and Microsoft 365. Under his leadership StepFun built the Step model family and raised over 5 billion yuan (about $700M) in a Series B+ round.
No articles ingested yet for StepFun. Once the hourly news pipeline is live, every article the classifier tags as mentioning this company appears here with its one-line AI summary and sentiment.
StepFun's series of multimodal foundation models, including the trillion-parameter-class Step-2 and later releases such as Step-3, with some variants (e.g. Step-3.5-Flash) distributed under open-source Apache 2.0 licensing.
A multimodal vision-language model in the Step line capable of processing image and text inputs together.
Multimodal models targeting speech and video understanding and generation, reflecting StepFun's focus on processing text, image, video and audio in one model line.
