Hacker News Reader

It's exceptionally optimized... 890 bytes per token. At standard 32K context, that's less than 30MB.

1. https://huggingface.co/deepseek-ai/DeepSeek-V4.1-Flash

2. https://www.mindstudio.ai/blog/deepseek-v4-1-flash-specs-arc...

Sure, but that size/token is inherent to the model. Part of DeepSeek's hosting advantage, and thus a remaining part of their moat, is migrating KV cache from VRAM to host RAM to nonvolatile storage, and presumably routing requests to where KV cache has migrated. People here on HN say KV cache lifetimes can be 8+ hours with DeepSeek, and I don't know of any other provider that can match that. That infrastructure isn't part of this open-source release.

That's right.

Yes, the theoretical approach is listed there. But the actual code is what's missing.

Reply on news.ycombinator.com