Chinese AI startup DeepSeek is reported to be launching V4 in mid-February, with internal testing claiming superior performance versus Anthropic Claude and OpenAI GPT models on coding tasks. The model would extend DeepSeek's market presence following attention to V3 and R1 releases. For financial institutions, this signals continued pressure on frontier model differentiation in a specific capability segment and reinforces the narrative of rapid Chinese capability catch-up in open development channels.
221 other tracked developments touching the same claims.