Use AiWings in your AI agent

AiWings adds image and video generation to your AI agent. You describe what you want in plain words, the agent calls AiWings, and the result appears in the conversation. Everything you make is also saved in your library.

• Images: GPT Image 2.5, 2 and 1.5, Nano Banana, Nano Banana 2 and Pro, Imagen 4 (Fast, standard and Ultra), Seedream 5 Lite, Grok Imagine and Flux Fast.
• Videos with sound: Seedance 2.0 and 2.5, from text, from an image, or guided by reference images and clips.
• Edits: change a result, combine your own photos, or animate an image into a video, all in the same chat.

You need two things:

• An AiWings account. Sign in with Google. You can also create it during the connect step.
• The AiWings server URL. Every agent uses the same one:

Pick your agent. Every agent uses the same server URL and your AiWings sign-in.

ClaudeChatGPTCursorAny MCP client

Works on claude.ai, the Claude desktop app and Claude mobile (once added on the web or desktop).

1. 1

   Open the custom connector window

   The button opens Claude with the window already up. Or go to Settings → Connectors and click Add custom connector.

   Open Claude connectors
2. 2

   Name it AiWings and paste the server URL

   Leave the advanced settings empty, then click Add.

   https://aiwings.io/api/mcpCopy
3. 3

   Connect and sign in

   Click Connect next to AiWings. A browser window opens: sign in to AiWings and approve access. You're sent back to Claude.
4. 4

   Turn it on in a chat

   In a new chat, open the tools menu and make sure AiWings is switched on. Then ask for an image or video.

On a Claude Team or Enterprise plan, an owner may need to add the connector for the organization first, in the organization's connector settings. Members then connect it from their own Settings → Connectors.

Ask your agent:

If it answers with your balance, you're connected. If it says it has no such tool, make sure AiWings is switched on for the chat, or reconnect it (see Troubleshooting).

Talk to it the way you'd brief a designer. You rarely need to name models or settings.

Three ways to start from your own images or footage:

Videos take about 1 to 5 minutes. The agent starts the job and a player in the chat updates until it's ready. You can leave the chat: finished videos are in your library. If a video fails, its credits are refunded automatically.

• Sound is on by default. Put spoken lines in double quotes in your prompt.
• First and last frames can't be combined with reference images. Use one or the other.
• Reference videos guide motion and camera, and use a higher per-second price.

Save the model and settings you use most so you don't have to repeat them. Set them on the Defaults page, from the result card in the chat (“Change defaults”), or just tell your agent:

A one-off request like “this time in 9:16” doesn't change your defaults.

Your agent picks the right tool from your request. These are listed for reference.

Two internal tools used by the upload box also appear in the list. They aren't meant to be called directly.

Images have a fixed price per quality level. Videos are billed per second. Credits per image:

Credits per second of video:

Plans: Basic $9/month for 1,000 credits; Pro $19/month for 2,000 credits; Scale $39/month for 4,000 credits. Top up anytime (100 credits per $1, from $5) on the billing page. Failed generations are refunded automatically.

Remove AiWings in your agent's connector or plugin settings. To cut off access from our side too, open your dashboard and click Revoke access under Connected apps. You can reconnect anytime.

Your uploads and results are stored by AiWings and available through your account. Prompts and reference files are only used to make your result. See the Privacy Policy.

Still stuck? See Support or email Mejed@templyo.io.