DeepSeek Releases V4 Open-Source AI Model, Claims Competitive with OpenAI and Google DeepMind
Pro version boasts 1.6 trillion parameters, both models feature 1M token context window
L'essentiel
- DeepSeek has released its next-generation V4 open-source AI model in two versions: V4-pro with 1.6 trillion parameters and V4-flash with 284 billion parameters.
- Both models feature a 1 million token context window, achieved with 'world-leading' cost efficiency, positioning the Hangzhou startup as competitive with leading US closed-source models.
Résumé généré par IA
Pourquoi c'est important
DeepSeek is a Hangzhou-based AI startup that has been gaining attention for its open-source AI models. The company has previously released competitive alternatives to leading US AI models.
DeepSeek has finally released its much-anticipated next-generation open-source foundational AI model, V4, which it said was competitive with leading US closed-source models from the likes of OpenAI and Google DeepMind. DeepSeek released two versions of the model on Friday, with the V4-pro model boasting 1.6 trillion parameters, making it the Hangzhou start-up's biggest ever model, while the smaller V4-flash model has 284 billion parameters. Both models have a context window of 1 million tokens, a critical feature that determines the amount of information an artificial intelligence system is able to process, which DeepSeek said was achieved with "world-leading" cost efficiency. More to follow...
Questions ouvertes
- What specific benchmarks did DeepSeek use to compare V4 with OpenAI and Google DeepMind models?
- What is the actual inference cost and performance of the V4 models?
- When will the models be publicly available for testing?
