AI 导读
💾 更小的 KV cache。更大的节省。 相比上一代,V4.1-Flash 的 KV cache 只需: 🔹 1/4 的 HBM 🔹 1/8 的 SSD 存储 缓存命中费用往往占智能体成本的很大一部分。压缩缓存能显著降低这些成本。 3/6
正文
💾 Smaller KV cache. Bigger savings.
Compared with the previous generation, V4.1-Flash’s KV cache needs just:
🔹 1/4 the HBM
🔹 1/8 the SSD storage
Cache-hit charges often account for a large share of agent costs. Compressing the cache cuts those costs significantly.
3/6
来源:@deepseek_ai · x.com