Gemini agentic video is live in the API and Claude still can't watch
Read the Portuguese translation →
Author: Jack Roberts · Watch on YouTube
Video authored by the channel shown, not by IANIA; all rights belong to the creator. Transcript and summary generated automatically; the original-language content is preserved.
Original language: English
Original video: Gemini agentic video is live in the API and Claude still can’t watch · Channel: Jack Roberts · 00:00 · Watch on YouTube
Topics: AI agents, AI developer tools, AI models, Multimodal models
Summary
Section titled “Summary”Claude cannot watch video directly and only reads its transcript, leaving on-screen content unseen. The speaker says Gemini’s agentic video feature addresses this and is now available in the API. They claim it uses up to 88% fewer tokens, cuts costs by up to 66%, and is more accurate. The video ends with a request to subscribe for more AI content.
Key points
Section titled “Key points”- The speaker says Claude only reads video transcripts and cannot see on-screen content.
- Gemini’s agentic video is described as a solution to this limitation.
- The feature is said to be live in the API.
- The speaker says it does not simply chop video into frames and read them.
- The speaker claims it uses up to 88% fewer tokens.
- The speaker claims it can cut costs by up to 66%.
- The speaker says it is more accurate.
- Viewers are asked to subscribe for more AI content.
Chapters
Section titled “Chapters”| Start | Chapter | Summary |
|---|---|---|
| 00:00 | Claude’s video limitation | The speaker says Claude reads transcripts but cannot see what appears on screen. |
| 00:00 | Gemini agentic video | The speaker introduces Gemini’s agentic video feature and says it is live in the API. |
| 00:00 | Efficiency claims and subscribe request | The speaker gives token-use, cost, and accuracy claims, then asks viewers to subscribe. |
Quotes
Section titled “Quotes”Claude can’t watch video. — 00:00
And it’s live in the API right now. — 00:00
It uses up to 88% fewer tokens — 00:00
cuts your cost by up to 66% — 00:00
Concepts and entities
Section titled “Concepts and entities”Claude, Gemini, agentic video, API
Full transcript (original language)
Section titled “Full transcript (original language)”00:00 Claude can’t watch video. It only reads the transcript. So, everything on screen is invisible to it. But, Gemini just fixed that with agentic video. And it’s live in the API right now. And it doesn’t just chop the video into frames and read it. It uses up to 88% fewer tokens, [music] cuts your cost by up to 66%, and it’s just much more accurate. [music] And if you want to stay at the cutting edge of AI, hit subscribe for more content like this.