One image in 4 seconds, cost of 1,000 images only $0.034, Google releases ultra-lightweight image model and opens video model Omni Flash for public beta

robot
Abstract generation in progress

According to Beating monitoring, Google has released two new multimedia generation models. Among them, Nano Banana 2 Lite (API name: gemini-3.1-flash-lite-image), which targets image generation, focuses on ultra-low latency: generating a single 1K image takes about 4 seconds, and the cost per 1,000 images is only $0.034—making it a high-speed, low-cost alternative to the previous-generation gemini-2.5-flash-image.

In addition, Gemini Omni Flash, a video model that debuted at Google I/O 2026, has opened a public beta for developers for the first time. The model allows users to edit videos through natural-language chat (such as adding digital magic effects or reworking spatial design), but the official note specifically points out that the current version still has limitations in character consistency when changing scenes or panning the camera. The model is now in public beta on the Gemini API and Google AI Studio (it had previously launched on the consumer-facing Gemini app and Google Flow).

View Original
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.
  • Reward
  • Comment
  • Repost
  • Share
Comment
Add a comment
Add a comment
No comments
  • Pinned