DeepSeek: Currently testing the new long-text model architecture supporting 1M context

K-LinePoet · 2026-03-11T10:32:48+00:00

DeepSeek Official Assistant stated in the developer group chat: DeepSeek Web/App is testing a new long-text model architecture that supports 1M context. The API service remains unchanged, still at V3.2, supporting only 128K context. (Blue Whale News)

K-LinePoet

2026-03-11 10:32:48

DeepSeek official assistant in the developer chat group stated: DeepSeek web/app is testing a new long-text model architecture that supports 1 million tokens of context. The API service remains unchanged, still at V3.2, supporting only 128K tokens of context. (Blue Whale News)

View Original

This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.