STORY · MODELLER_

DeepSeek-V3 reaches a million tokens – and agents can actually use it

DeepSeek has launched a model that supports up to a million tokens in context window, a significant increase from earlier models. With this capacity, AI agents can handle substantially larger and more complex tasks without losing information along the way.

WHY IT MATTERS

A million-token context is a shift for practical use of AI agents – it enables analysis of entire codebases, long document sequences, and complex multi-step tasks that were previously technically impossible. This challenges competitors who must improve their context window sizes to keep pace.

SOURCES

MACHINE-GENERATED SUMMARY This summary is written by machine from the sources below. We sort and explain — but we are a way into the field, not the final word. Check the source when something matters to you.