Analysis shows lower cost but slower and less consistent results at scale.
The New Stack reports that DeepSeek has released its first vision model. When compared directly to Gemini 3.7 Flash the new model delivered matching performance across nine vision questions. It achieved this result at roughly one third the price according to their testing. However the responses arrived more slowly and with greater variability in quality. These characteristics become significant when applications must process queries at high volume. The piece frames the decision as a choice between lower spending and faster more reliable output for development teams.
556
1 post, first seen 22h ago
Analysis shows lower cost but slower and less consistent results at scale.
The New Stack reports that DeepSeek has released its first vision model. When compared directly to Gemini 3.7 Flash the new model delivered matching performance across nine vision questions. It achieved this result at roughly one third the price according to their testing. However the responses arrived more slowly and with greater variability in quality. These characteristics become significant when applications must process queries at high volume. The piece frames the decision as a choice between lower spending and faster more reliable output for development teams.
The New Stack@thenewstackDeepSeek's first vision model vs. Gemini 3.7 Flash: It comes down to spend vs. speed
The New Stack@thenewstackDeepSeek's first vision model vs. Gemini 3.7 Flash: It comes down to spend vs. speed
