The agent uses Unsplash's search endpoint with the customer's keywords, orientation preferences, and color filters. Results come back with thumbnails, full-resolution URLs, photographer info, and download links, all presented conversationally to the customer.

Yes. Unsplash requires proper attribution per their API guidelines. The agent automatically includes photographer name and Unsplash credit with every photo result, making compliance effortless for your team.

Unsplash allows 50 requests per hour in demo mode and 5,000 per hour with production approval. For most customer-facing agents, production access is recommended. Apply through the Unsplash developer portal to unlock higher limits.

The agent filters by orientation (landscape, portrait, squarish) and retrieves images at multiple resolutions. Unsplash provides raw, full, regular, small, and thumb sizes for every photo, so the agent serves the right resolution for each use case.

No. Tars queries the Unsplash API in real time and returns URLs to the images. Photos are served directly from Unsplash's CDN. No images are downloaded to or cached on Tars servers.

Instead of leaving your conversation to browse a separate website, your AI agent brings the entire Unsplash library into the chat. Customers describe what they need in natural language and get curated results instantly, without navigating search interfaces or switching tabs.

Yes. If you have curated Unsplash collections for your brand or project, the agent can search within specific collection IDs. This narrows results to pre-approved visual styles that match your brand guidelines.

The agent handles empty results gracefully by suggesting alternative keywords, broadening the search criteria, or offering to search with different filters. Customers always receive helpful next steps rather than a dead end.