Skip to main content


DeepSeek has unveiled two preview versions of its newest large language model, DeepSeek V4. The Pro model has 1.6 trillion parameters (49 billion active), making it the biggest open-weight model available. Both models have almost closed the gap with frontier models on reasoning benchmarks, with V4-Pro-Max outperforming GPT-5.2 and Gemini 3.0 Pro on some tasks. They also undercut competitors on pricing, with V4 Flash costing 0.14 USD per million input tokens. techcrunch.com/2026/04/24/deep… #AIagent #AI #GenAI #AIResearch