{"data":{"id":1529,"lang":"en","title":"DeepSeek Unveils V4 Open-Source Flagship With 1 Million Token Context, Claims Performance Near GPT-5.4 and Gemini 3.1-Pro","summary":"DeepSeek released open-source V4-Pro and V4-Flash models with a 1M-token context window, claiming performance close to GPT-5.4 and Gemini 3.1-Pro.","category":"technology","category_label":"Technology","published_at":"2026-04-24T14:01:37Z","updated_at":"2026-04-24T14:01:37Z","url":"https://techtools.cz/tools/news/article/1529-deepseek-unveils-v4-open-source-flagship-with-1-million-token-context-claims","api_url":"https://techtools.cz/news-api/v1/articles/1529?lang=en","markdown_url":"https://techtools.cz/news-api/v1/articles/1529.md?lang=en","image_url":"https://hongkongfp.com/wp-content/uploads/2025/01/DSC06277-Copy-1050x700.jpg","word_count":160,"reading_time_minutes":1,"sources":[{"url":"https://www.cnbc.com/2026/04/24/deepseek-v4-llm-preview-open-source-ai-competition-china.html","domain":"cnbc.com"},{"url":"https://hongkongfp.com/2026/04/24/chinas-deepseek-releases-long-awaited-new-ai-model/","domain":"hongkongfp.com"},{"url":"https://www.aljazeera.com/economy/2026/4/24/chinas-deepseek-unveils-latest-model-a-year-after-upending-global-tech","domain":"aljazeera.com"}],"translation":{"lang":"cs","id":1602,"title":"DeepSeek představil open-source vlajkový model V4 s milionovým kontextem, výkon se blíží GPT-5.4 a Gemini 3.1-Pro","url":"https://techtools.cz/tools/news/clanek/1602-deepseek-predstavil-open-source-vlajkovy-model-v4-s-milionovym-kontextem-vykon","api_url":"https://techtools.cz/news-api/v1/articles/1602?lang=cs"},"content":"Chinese AI startup DeepSeek released preview versions of its long-awaited V4 flagship on April 24, 2026, arriving just over a year after the company shocked Silicon Valley with its original breakthrough. The lineup includes V4-Pro with 1.6 trillion parameters and V4-Flash with 284 billion parameters, both open source and featuring a new Hybrid Attention Architecture.\n\nThe standout feature is a 1 million-token context window, large enough to fit entire codebases or book-length documents in a single prompt. DeepSeek claims V4-Pro outperforms Anthropic Claude Sonnet 4.5 and approaches Opus 4.5, falling only marginally short of OpenAI GPT-5.4 and Google Gemini 3.1-Pro. The company acknowledged its trajectory trails frontier Western models by roughly three to six months.\n\nThe release ratchets up competition in the open-source AI race, pressuring US labs that keep their top models proprietary. It arrives the same week OpenAI launched GPT-5.5 and Microsoft integrated Anthropic models into its security stack, underscoring how quickly the frontier is moving in 2026."}}