
Veo
Your AI agent uses Google Veo 3 to generate 8-second 720p videos with native audio directly inside live conversations. A customer describes a scene, the agent submits the prompt, monitors rendering progress, and returns the finished video. Text-to-video production happens mid-chat without any manual steps.




Your AI agent turns text prompts into polished video clips through Veo's generative models, managing every step from prompt submission to file download in real time.
Veo
See how creative teams and marketers use AI agents with Veo to produce video content, prototype visual concepts, and generate social media clips through conversation.
A marketing manager messages 'create a 9:16 video showing a coffee cup on a wooden table with morning sunlight.' Your AI Agent sends the prompt to Veo 3 with the vertical aspect ratio, monitors the rendering operation, and delivers the finished 8-second clip with ambient audio. The manager downloads it and posts directly to Instagram Stories without involving a video production team.
A creative director wants to compare two visual directions for an ad campaign. Your AI Agent generates one video with a 'cinematic slow motion' prompt and another with an 'upbeat fast cuts' prompt using the fast model version. Both clips render within minutes, giving the director two options to evaluate before committing to full production.
An e-commerce team needs a short video showing their new sneaker in motion. Your AI Agent takes the text description, applies a negative prompt to exclude competitor branding, selects the high-quality veo-3.0-generate-preview model, and returns a 720p clip. The team uses the generated video as a placeholder in their product listing while the professional shoot is still being scheduled.

Veo
FAQs
Veo 3 generates 8-second videos at 720p resolution with natively produced audio. The clips are ready for social media, product pages, or creative review. Higher resolution and longer durations depend on the model version Google makes available through the Gemini API.
The agent supports all Veo models available through the Gemini API, including veo-3.0-generate-preview for highest quality and veo-3.0-fast-generate-preview for faster generation. The agent can also list all available models so you can choose the best fit for your use case.
Generation time depends on the model version and current API load. The fast model typically renders in under a minute, while the standard model may take several minutes. The agent polls the operation status automatically and notifies you as soon as the video is ready.
Yes. The agent passes your preferred aspect ratio directly to Veo. Supported ratios include 16:9 for landscape, 9:16 for vertical mobile content, and 1:1 for square social media posts. Specify the ratio in your prompt or let the agent ask.
No. Tars sends prompts to Google's Veo API and retrieves the download URI from the operation response. The video file is hosted by Google. Tars provides the download link within the conversation but does not cache or store video files on its servers.
You provide a Gemini API key from Google AI Studio. This key is sent as the x-goog-api-key header with every request. No OAuth flow or multi-step setup is needed. The same key works for all Veo model versions.
Yes. Beyond the main text prompt, the agent supports negative prompts to exclude specific visual elements. You can also enable or disable person generation depending on regional constraints and content policies. These parameters give you fine-grained creative control.
Google AI Studio requires manual interaction through a web interface. The Tars integration embeds Veo into conversational channels like Slack, WhatsApp, or your website chat. Teams generate videos by describing what they need in natural language without switching to a separate tool.
Don't limit your AI Agent to basic conversations. Watch how to configure and add powerful tools making your agent smarter and more functional.

Privacy & Security
At Tars, we take privacy and security very seriously. We are compliant with GDPR, ISO, SOC 2, and HIPAA.