DeepSeek 的实验性视觉模型支持三种格式，将图像限制为 384 个标记
Liang Wenfeng 单独命名的 deepseek-v4-flash-vision-exp 支持 Chat Completions、Messages 和 Responses 请求，每张图片计费不超过 384 个 V4-Flash 令牌。
By RuntimeWire Staff · Published
Primary source: DeepSeek
为何重要
DeepSeek 已为一个单独命名的实验性模型附加了一个可预测的图像输入上限，该模型通过三种熟悉的请求格式提供。生产环境的 V4-Flash 端点仍然仅支持文本，并且 DeepSeek 未发布任何用于评估该视觉模型的基准、延迟或可靠性数据。
Liang Wenfeng's DeepSeek announced deepseek-v4-flash-vision-exp on August 21st, 2026, an experimental model for mixed text and image requests. The announcement names three compatible request formats - Chat Completions, Messages and Responses - and caps each image at 384 billable tokens under V4-Flash pricing.
When you purchase through links in our articles, we may earn a small commission. This doesn’t affect our editorial independence.