Coverage (30d)
3vs0
This Week
1vs0
Evidence
0 articlesRelationships
0Timeline
DeepSeek V42026-07-07
Research paper published introducing FlashMemory-DeepSeek-V4 with Lookahead Sparse Attention for ultra-long context
DeepSeek V42026-06-09
DeepSeek V4 technical disclosure: 500K context with 90% less KV cache using FlashMemory
DeepSeek V42026-05-22
DeepSeek v4 API pricing permanently cut 75% to $0.43/M input tokens and $0.87/M output tokens
DeepSeek V42026-05-10
DeepSeek v4 launched, exposing performance gap in AMD ROCm vs NVIDIA CUDA
DeepSeek V42026-04-13
Deployment of DeepSeek V4 on Huawei's Ascend AI chips discussed as part of China's semiconductor counterstrategy.
DeepSeek V42026-04-07
Began limited gray-scale testing with a new tiered interface (Fast, Expert, Vision modes).
DeepSeek-V32024-12-01
Released in December 2024, demonstrating Mixture-of-Experts architecture with 671B parameters
DeepSeek-V32024-12-01
Released DeepSeek-V3, a 671-billion-parameter model
Ecosystem
DeepSeek-V3
deploysMixture of Experts (Sparse MoE for LLMs)1 src
DeepSeek V4
usesFlashMemory1 src
competes withGPT-4o1 src
competes withClaude 3.5 Sonnet1 src
competes withGemini 3 Pro1 src
usesMixture of Experts1 src
Benchmarks
mmlu pro
DeepSeek-V385
DeepSeek V4—
arena elo
DeepSeek-V31380
DeepSeek V4—