Gemini
Use Gemini through AI to generate text, create images, and retrieve finished Veo videos with your API key.
This connection provides Gemini text and speech generation, embeddings, token counting, image generation, and asynchronous Veo video generation.
What your agent can do in Gemini
Actions grouped by task. Each group lists the actions it calls.
Text and Embeddings
Generate text or speech, embed supplied text, and count input tokens.
- Generate Content
- Embed Content
- Count Tokens
Images and Video
Generate images, submit Veo video jobs, and retrieve their status or finished files.
- Generate Image
- Generate Videos
- Get Videos Operation
- Wait For Video
Tasks to hand your agent
Fill in the brackets, and connect any second app a task names.
Draft Text from a Brief
Generate a draft using your source material and tone requirements.
Use Gemini model [model ID] to write [document type] from these notes [notes], using [tone] and no more than [length].
Generate and Retrieve a Video
Start a Veo generation and wait for the finished video file.
Generate a Veo video using model [model ID] with this scene: [scene description]. Use [aspect ratio] and [duration]. Wait for completion and give me the video link.
Actions8
Every action the Gemini connector offers, by name and id.
- List Models
list_modelsList the available Gemini and Veo models. - Generate Content
generate_contentGenerate text or speech audio with Gemini models. - Embed Content
embed_contentGenerate Gemini embeddings for text content. - Count Tokens
count_tokensCount the Gemini token usage for input text. - Generate Image
generate_imageGenerate an image with Gemini and return a transit URL. - Generate Videos
generate_videosStart a Gemini Veo video generation operation. - Get Videos Operation
get_videos_operationFetch the current status for a Gemini Veo operation. - Wait For Video
wait_for_videoPoll a Gemini Veo operation and return the finished video via transit URL.
How Gemini signs in
- Sign-in method
- API keyConnect a Gemini API key from Google AI Studio.
- You need
- A Gemini API key with access to the model you choose.
- Notes
- Image and finished-video delivery require the product’s file transit service. Video submission starts an operation; retrieve its result after processing.