deepseek-v4.1IMAGE
DeepSeek-V4.1-Flash-1017 hits over 150 TPS on Inference.net
A user reports the model running above 150 tokens per second on the Inference.net inference platform, with a screenshot.
Ibrahim Ahmed
@atbemeArchived 11 days agoOriginally posted 2026.10.18
Ibrahim Ahmed @atbemeOriginally posted 2026.10.18
speed speed speeeeeeeed DeepSeek-V4.1-Flash-1017 is running over 150 TPS on @inference_net
This post quotes another post: x.com/deepseek_ai/status/2111375640717033777
x.com/atbeme/status/2111545525929759175Archived 2026.10.18 16:29