DeepSeek V4.1 Flash hits 1,521 tok/s on Fireworks Ultrafast, ships Tetris in 11 seconds
nicolechirps · x · 2026-10-22
Fireworks AI says DeepSeek V4.1 Flash 1017 runs at 1,521+ tok/s on its Ultrafast inference stack — fast enough to generate a working Tetris game in 11 seconds, claimed 7x faster than Claude Opus 5.5. Ultrafast is currently waitlist-only.
More from Infra
- SemiAnalysis' GB300 vs DeepSeek profit chart called most misleading of the year: real gap is 10.9x — teortaxesTex · 2026-10-22
- H100 Rental Rates Keep Climbing Across the Entire Term Structure — isidentical · 2026-10-22
- Phonon TTS and Fluid Speech Mini: a day-long test of lightweight local speech models — gregbarbosa · 2026-10-22
- Databricks AI Runtime goes GA: GPUs in seconds, adopted by Robinhood, Superhuman, Merck — matei_zaharia · 2026-10-22
- Flint MoE model lands in llama.cpp: 39.8B total params, only ~2.5B active per token — jacek2023 · 2026-10-22
- Omarchy package server is timing out: updates failing on cursor, grok-bot, codex-desktop packages — mark_k · 2026-10-22