DeepSeek
Maker · 11 entries
Everything from DeepSeek
| Name | What to know | Price or memory | |
|---|---|---|---|
| strong Best-value open-weights model: near-frontier work for a few dollars, though a step behind. | ~510 GB | ||
| DeepSeek app features exposed via API: search, files, OCR, ASR, TTS, with separate tool fees and an open-source frontend | |||
| promising Free, MIT-licensed coding harness that ties for first on a leaderboard, with one scoring flaw. | |||
| unproven DeepSeek's serving engine is open, but its production economics can't be reproduced. | |||
| Until V4.1 Pro ships, all V4 Pro requests are routed to V4.1 Flash and billed at Flash prices | ~893 GB | ||
| Full, reindex and reuse layer modes share KV cache and indices across layers | |||
| Multi-token-per-step decoding component included in V4.1 Flash | |||
| Separate static-fact memory module held in server RAM; 168B params per video (196B per HF page) | |||
| Fuses operations to cut intermediate GPU memory traffic | |||
| Discards local sliding-window cache after each reply and recomputes the last 128 tokens on the next turn | |||
| Experimental vision version of V4 Flash, open-sourced on Hugging Face after earlier API release; 305B parameters, about 168 GB | ~168 GB | ||