On September 5th, Jinshi Data reported that Tencent officially released a new generation of large-scale model - Tencent Huan Yuan Turbo at the 2024 Tencent Global Digital Ecological Conference. The model adopts the MoE architecture, which improves the inference efficiency by 100% compared to the previous generation product, and reduces the inference cost by 50%. Externally, the price of Tencent Huan Yuan Turbo is also 50% lower than that of Huan Yuan Pro, with an output price of 0.05 yuan/thousand tokens and an input price of 0.015 yuan/thousand tokens.
This page may contain third-party content, which is provided for information purposes only (not representations/warranties) and should not be considered as an endorsement of its views by Gate, nor as financial or professional advice. See Disclaimer for details.
1 Likes
Reward
1
3
Repost
Share
Comment
0/400
Markman
· 2024-09-08 12:04
WAGMI 💪
Reply0
GateUser-cb09837c
· 2024-09-05 03:25
they are in the way you are doing the same thing as a meeting with the kids and I can get it on the way to work and I can get it on the way to work and I can get it on your way to get a meeting with the kids and you are in a meeting with the girls in a good place and you can do it now and you can do it and you are doing it now and you can
Tencent releases new generation large-scale model 'Mix Yuan Turbo', reduces inference cost by 50%
On September 5th, Jinshi Data reported that Tencent officially released a new generation of large-scale model - Tencent Huan Yuan Turbo at the 2024 Tencent Global Digital Ecological Conference. The model adopts the MoE architecture, which improves the inference efficiency by 100% compared to the previous generation product, and reduces the inference cost by 50%. Externally, the price of Tencent Huan Yuan Turbo is also 50% lower than that of Huan Yuan Pro, with an output price of 0.05 yuan/thousand tokens and an input price of 0.015 yuan/thousand tokens.