LINKED LIST [txt mode] ▸ DeepSeek V4.1 Flash 1017 (max) - Inte...
home explore | log in

DeepSeek V4.1 Flash 1017 (max) - Intelligence, Performance & Price Analysis

artificialanalysis.ai · first added by @jdorsey · 2026-10-24 · 1 upvotes

log in to save, upvote or flag this.


─── In 0 lists ─────────────────────────────────────────

(not in any lists yet)


─── Discussions ────────────────────────────────────────

* DeepSeek V4.1 Flash 1017 (Max)
190 pts · 96 comments · node

─── From the discussion ────────────────────────────────

* Run Qwen 3.8 Flash Next (125B) on consumer hardware (RTX 4090) at 100T/s
github.com · node
* lxe/dsx: DeepSeek-V4.1-Flash-1017 (552B MoE) on 2x RTX 3090 + 128 GB RAM: Rust + CUDA engine and 2-bit quant. 30-37 tok/s decode.
github.com · node
* The plunging price of thought
epoch.ai · node
* Deepseek V4.1 Flash 1017 Max on Three.js
threejseval.com · node
* DeepSeek V4.1 Flash 1017 API - Demo
deepinfra.com · node