DeepSeek V4.1 Flash: The Native Multimodal Model That's Breaking Speed Records ...
DeepSeek V4.1 Flash: The Native Multimodal Model That's Breaking Speed Records
How DeepSeek Achieved 420 Tokens/Second Without Sacrificing Accuracy
Published: September 9, 2026 | Reading time: 10 minutes
The Surprise Beta
On September 8, 2026, DeepSeek quietly launched a beta test for V4.1 Flash, an intermediate model with the identifier deepseek-v4.1-flash-expires-on-0910. The model name itself reveals the urgency—this beta expires on September 10, giving developers just 48 hours to test it.
DeepSeek begins limited-time beta of V4.1 Flash multimodal model · TechNode
DeepSeek: China's DeepSeek launches V4.1-Flash model - The Economic Times
New Deepseek model V4.1-Flash cuts memory needs for AI agents
DeepSeek-V4.1-Flash enters testing phase, challenges AI model rankings
DeepSeek releases V4.1-Flash, says it outperforms flagship V4-Pro - SiliconANGLE
DeepSeek formally launches V4.1 Flash, routes V4 Pro requests to Flash · TechNode
DeepSeek launches V4.1-Flash model with 552B parameters and a million-token context window
DeepSeek says new Flash AI model beats Kimi K3 on cyber, coding benchmarks
China’s DeepSeek Launches Smaller, Faster AI Model
DeepSeek's V4.1-Flash takes on Opus 5, and retires V4-Pro
