DeepSeek V4 Flash 0731: Frontier Coding at $0.14/M
DeepSeek V4 Flash 0731 hits 82.7 on Terminal-Bench at $0.14/$0.28 per million tokens, beating its own Pro model. Full be…
Chinese AI Models Now Power Up to 46% of Enterprise API Traffic — Here's Why (2026)
Chinese open models like DeepSeek and Qwen now drive up to 46% of enterprise API traffic on OpenRouter. See the real num…
DeepSeek V4 Explained: How Hybrid Sparse Attention Cracked the 1 Million Token Context in 2026
DeepSeek V4 ships 1.6T params, 1M-token context, 80.6% on SWE-bench under MIT. Inside the hybrid attention that cut KV c…