DeepSeek
Also known as: DeepSeek AI, DeepSeek-V4, DeepSeek R1
DeepSeek is an AI company based in Hangzhou, China, spun out of the hedge fund High-Flyer. It releases its language models as open-source LLMs, free to download and use commercially. In early 2025, its reasoning model R1 made global headlines: frontier-level performance at a fraction of typical training costs.
Current models
DeepSeek-V4 has been available since April 2026, in two variants: V4-Pro (1.6 trillion parameters, 49 billion active) and V4-Flash (284 billion parameters, 13 billion active). Both use a mixture-of-experts architecture and handle up to 1 million tokens of context. V4.1-Flash followed on 10.09.2026: multimodal, 552 billion parameters in a mixture-of-experts architecture, also up to 1 million tokens of context. Requests to V4-Pro have been routed to V4.1-Flash since 14.09.2026, until V4.1-Pro launches.
Benchmark standing
On agentic benchmarks, V4-Pro scores at the level of leading Western models from OpenAI and Anthropic, at significantly lower API prices.
Open source and self-hosting
The V4 models, including V4.1-Flash, are released under the MIT license. Companies can download the weights, adapt them, and run them as self-hosted AI on their own infrastructure.
Data protection in the China context
For EU companies, one distinction matters: the official DeepSeek API and app process data on servers in China. Self-hosting avoids this entirely: the model weights themselves transmit no data. See digital sovereignty.
We can work out the next step together in a free introductory call. Contact →
Last updated: October 2026
Claude Cheat Sheet
An overview of Claude's features and models, with tips from our daily work. It is available in German and English, and we keep it up to date.
Open the cheat sheetDigital Impact in your inbox
With one sign-up you get three newsletters: the AI Insights Newsletter with our new articles every week, plus the Digital Impact Longread and the Digital Impact Update once a month. Sign-up uses double opt-in, and you can unsubscribe at any time.
AI Operations as a podcast
In the Digital Impact Podcast, André Hellmann, Christina D'Ilio, Christian Sattel and Sarah Stock talk with guests from practice about AI Operations. The episodes are easy to listen to on the go.
How this article was produced
- Topic selection
- Source selection
- Fact-checking
- Approval
- Research
- Drafting
- Diagrams
- Publishing
This article was produced with AI support. Ideation, editorial planning, substantive review and approval rest with a human; copy-editing sits with the AI. Editorial responsibility is held by André Hellmann.