Budget: 14900 UAH Deadline: 10 days
Hello! I checked the request regarding the automation of video generation in Flow (Veo 3.1 Fast). An important point: the web interface of Flow on labs.google is protected by reCAPTCHA and subscription/regional restrictions. Performing "request emulation" to the web interface with automatic bypass of reCAPTCHA is unstable (it may break after any changes in the frontend) and may violate the service's terms of use. Instead, there is a correct and supported way: official integration of Veo through Google Cloud (Vertex AI). There are scenarios for text-to-video, image-to-video, and generation from first/last frames, and this can be implemented purely through HTTP requests (without Selenium), with normal quotas/logs/retries. If the official option through Vertex AI suits you, I propose the following plan: 1) We agree on the format of input data (image / 2 frames / prompt), required parameters (duration, resolution, video format). 2) You provide access/settings for Google Cloud: Project + Billing, enabled Vertex AI, Service Account (key/permissions) + region/quotas. 3) I will implement a Python script/CLI: it accepts image+prompt, sends a request to Vertex AI, waits for job completion, saves the video, maintains a log, handles errors/retries. Estimated timeline: 5–10 working days after receiving access.