Connect once with OAuth, unlock 6 bonus credits, then ask Claude for an image or video in any conversation.
Works in Claude Code, claude.ai, and Claude Desktop using the same balance as the web generators.
Connect → unlock → create
Connect once, receive the one-time credit reward, then copy a real task into Claude. No API key and no prompt handoff to a separate generator.
Choose your Claude client
Name it “Claude Imagine” and paste this URL:
Your first authenticated connection adds a one-time 6-credit reward, valid for 14 days. Together with 9 signup credits, that is enough for the guided image-to-video demo below. Need more later? Choose a subscription or a non-expiring credit pack.
Your first real task
The 9-credit workflow deliberately uses the full free trial: 3 credits for Nano Banana 2, then 6 for a four-second Seedance Pro clip at 480p.
Start small
Turn the context already in your Claude conversation into a polished website visual.
The full demo
Generate a key visual, then pass its returned public URL straight into the video tool.
Beyond a blank prompt box
The same models are available on the web. MCP changes the workflow: Claude carries the conversation context, can inspect a codebase, checks live costs, and hands one tool's output to the next.
Claude already knows the audience, positioning, and approved idea. It can turn that shared context into the visual brief instead of making you rewrite it in another tool.
“Review our conversation, summarize the approved visual direction in three bullets, then use Claude Imagine and Nano Banana 2 to create a 16:9 campaign key visual with room for a headline.”
Claude Code can inspect the page copy, theme variables, and asset folders before generating an OG image that belongs to the project it is already editing.
“Read the landing page copy and theme colors in this repository. Propose one OG-image concept, then use Claude Imagine to generate it at 16:9 and save the returned image under public/og/.”
The image tool returns a public URL. Claude can pass that exact result to the video tool in the same conversation, preserving your creative handoff.
“Use the public URL returned by the last Claude Imagine image and animate it into a 4-second 480p video with Seedance Pro. Preserve the subject, framing, and color palette; add only subtle camera movement.”
Before anything is charged, Claude can inspect your live balance and model prices, recommend the right trade-off, and wait for approval.
“Call get_credits and list_models. Recommend the best-value model for a 9:16 launch visual, explain the credit cost and trade-off, and wait for my approval before generating anything.”
WATCH THE WORKFLOW · 2 MINUTES
Follow along as we connect Claude Imagine, create a coffee campaign image, and bring it to life — all in one Claude conversation.
Connect Claude → Create an image → Animate it. English narration.
Watch on YouTubeOnce connected, Claude can call these directly. Images and videos use the same account balance.
Generate an image from a text prompt. Optional model (nano-banana-2 default, gpt-image-2.5, seedream-4.5, flux-2-pro, z-image) and aspect_ratio (1:1, 4:3, 3:4, 16:9, 9:16). Returns the image (an inline card in claude.ai, Claude Desktop and the mobile apps) plus its URL; failed generations are refunded. Optional style takes an id from list_styles.
Edits an existing image from a text instruction. Takes the image URL (normally the one a previous generation returned), an optional model (nano-banana-2 default, gpt-image-2.5) and an optional aspect_ratio; when you omit it the framing of the source image is kept.
Named looks for generate_image: YouTube thumbnails, book covers, album and podcast cover art, posters, blog covers and product photos, four each. In claude.ai they appear as a picker card; pick one and Claude generates in that style.
Generate from text or animate one public image with H3 Max Turbo, Grok Imagine, Seedance 1.5 Pro, Seedance 2.0 Mini, Veo 3.1 Fast, or Kling 2.5 Turbo. Choose supported seconds and quality; the server calculates the exact credits and refunds failed generations.
Returns a generation that was still running when its call ended. Claude calls it by itself with the task id quoted in that reply, so a slow render still lands in the same conversation.
Tells Claude the exact credit cost of a generation before it runs, with your balance and plan eligibility, so the price you hear is the price you pay.
Finds your own recent images and videos, including ones made on the website, so a new conversation can show, edit or animate an earlier result.
Lists the available image and video models and their live credit costs for your current plan tier.
Returns your remaining credit balance so Claude knows how much you have left before generating.
Plans a product-photo workflow end to end: two image directions from one original photo, your choice between them, then a short video. Returns an itemised quote at your prices, the prompts and an output checklist. Free, and it starts nothing.
Connect once, then continue with the video generator, the image-specific guide, or the complete MCP reference.
Start with 9 free credits, then unlock 6 more with your first authenticated MCP connection—up to 15 free credits in total. Videos start at 6 credits. Subscribe for recurring credits and perks, or buy a one-time pack that never expires.
Best for ongoing use. More credits per dollar than credit packs, refreshed every billing cycle, with more parallel generations. Cancel anytime.
Includes
Includes
Includes
Have another question? Email us at support@claudeimagine.com.
Client availability verified July 10, 2026 against Claude's custom connector documentation. Connector tools and authentication are verified against this site's MCP implementation.
