TechFlow news, on March 11, according to Jinshi Data, Silicon Flow announced that its SiliconCloud platform now supports batch inference for DeepSeek-R1 & V3 APIs. Starting immediately, users can send requests via the batch API to SiliconCloud without being constrained by real-time inference rate limits, with tasks expected to complete within 24 hours.
Compared to real-time inference, DeepSeek-V3 batch inference pricing has been reduced by 50%. From March 11 to March 18, DeepSeek-R1 batch inference offers an additional 75% discount, with input priced at 1 yuan per million tokens and output at 4 yuan per million tokens.
Add to Favorites
Share to Social Media
