DeepSeek V4.1-Flash hands-on: 350 t/s decoding speed but still very experimental
teortaxesTex · x · 2026-09-08
- Early tester ivanfioravanti reports DeepSeek V4.1-Flash averages 350 tokens/s decoding, impressively fast, though its architecture remains unclear.
- Issues so far: frequent failures under heavy load, excessive thinking, failing file edits, and 4M tokens without finishing a Frogger game.
- Counterpoint: teortaxesTex says the model sees images fine and can fix its own output if asked—play with it, but it's clearly early-stage.
Related event: DeepSeek V4.1 Flash Tested: Blazing 350 Tokens/s but Still Experimental(2 posts)→
More from Models
- Kimi K3.1 Pro tops MathArena open models, closing frontier gap to about 6 months — teortaxesTex · 2026-10-26
- Ruff creator Charlie Marsh: frontier models find infinite bugs in any software — charliermarsh · 2026-10-26
- Dev testing one skill across Opus 5.5 and Sol 6.1 finds instruction following 'all over the map' — lucasmeijer · 2026-10-26
- supermemory unveils learner-1, claiming continual learning for any model — julianweisser · 2026-10-26
- Anthropic says Opus 5.5 cuts cost per task by ~40% vs Opus 5 — taufiqintech · 2026-10-26
- AI search APIs ranked: top 12 within 7 points, yet prices differ nearly 7x — FinanceYF5 · 2026-10-26