Claude and 
ChatGPTArtificial Studio MCP connects 100+ AI models (Veo 3.1, Kling 3.0, Nano Banana Pro, GPT Image, ElevenLabs) to the chat you already use. One account, one credit balance, no tab-switching.
Click copy. You'll paste it in the next step.
The button opens Claude's Add custom connector window. Or go to Customize → Connectors → Add → Add custom connector.
Paste the URL, click Continue, then Add. Sign in to Artificial Studio, approve access, and ask for your first image.
No API key · 20 free credits every day · Set up in 2 minutes
My client is opening Crumb, a neighborhood sourdough bakery. Design the logo, show it on a tote bag and the storefront, and make a 10-second Instagram teaser with music.
On it. I’ll design the logo, place it on the tote and the storefront, then turn the storefront into a teaser that pushes in to the logo, and score it.
Logo
Tote bag
StorefrontDone. 203 credits for the full brand launch. Want a 9:16 story version with the opening hours?
The Artificial Studio MCP is a remote Model Context Protocol server that connects any AI assistants to 100+ AI models for images, video, audio, music and 3D. You describe what you need in the chat; your assistant picks the right model, shows you the cost, generates it and returns the result.
It’s built for designers, creative agencies and freelancers who already work with an AI assistant and want to produce finished creative assets without switching between separate AI tools.
100+
AI models for image, video, audio and 3D
20
Free credits every day to try any model
2 min
To connect to Claude or ChatGPT, no API key
Every tool in Artificial Studio is available from the chat: generation, editing, translation, upscaling and ready-made creative apps.
Brief in. Campaign out.
Copy a prompt, paste it into Claude or ChatGPT with Artificial Studio connected, and watch the assistant chain the tools for you.
“Create a launch campaign for my sneaker brand: 3 ad images in 1:1, 4:5 and 9:16 with Nano Banana, animate the best one into a 5-second vertical video with Kling 3.0 Standard, and add a 15-second upbeat track.”
≈ 164 credits (3 images × 8 + 5 s × 25 + 15 s × 1)
“Design a logo for Mono, a small ceramics studio, then show it on a tote bag, a coffee cup and a business card.”
≈ 32 credits (1 logo × 8 + 3 mockups × 8)
“Translate this 30-second product video into Spanish and French, keeping my voice.”
≈ 480 credits (30 s × 8 × 2 languages)
“Turn this photo of our chair into a 3D model with Trellis 2, then render it in a sunlit Scandinavian living room.”
≈ 70 credits (Trellis 2 × 60 + GPT Image 2 render × 10)
Separate AI tools make you the integration layer: copy a prompt here, download there, re-upload somewhere else. With the Artificial Studio MCP, your assistant does the switching.
Ask for a product image, animate it, add music and upscale it to 4K in one message. Claude or ChatGPT runs every step for you: no tabs, no downloads and re-uploads between tools.
Ideogram for text in images, Nano Banana for consistency, Kling Turbo when you're on a budget. Your assistant compares the options, tells you the credit cost and asks before it spends.
Your assistant already has your brief, your notes and your project files. It writes better prompts from that context, and in Claude Code or Claude Desktop it can upload your own images, videos and audio as inputs.
"Make 10 versions of this ad in 1:1, 4:5 and 9:16" is one sentence, not an afternoon. The repetitive production work is exactly what the assistant handles best.
The same account powers the MCP, the REST API and the n8n node, so a workflow you design in a chat can become an automation that runs on its own.
Deliver more. Hire less.
Agencies use the Artificial Studio MCP to produce ad images, videos, voiceovers and music for clients directly from Claude or ChatGPT. One subscription replaces separate seats for image, video and audio tools, and every teammate works from the same credit balance.
Say yes to more briefs.
Freelancers use the Artificial Studio MCP to offer video, music, voiceover and 3D work without learning new software or subcontracting. Go from brief to a first draft the same day, inside the assistant you already use to plan the project.
Skip the busywork, keep the craft.
Designers use the Artificial Studio MCP to generate concepts, mockups, vector graphics and 3D models, then refine them in their own tools. 3D models arrive as GLB files ready for Blender, and the Artificial Studio Blender extension lets you generate without leaving the viewport.
A selection of the 100+ models you can use from the chat, with their cost in Artificial Studio credits. Your assistant always shows the cost before generating.
Prices last updated
| Model | Developer | Cost in credits (per image) |
|---|---|---|
| Nano Banana Pro | 30 credits per image | |
| Nano Banana 2 | 16 credits per image | |
| Seedream 5.0 Pro | ByteDance | 14 credits per image |
| Ideogram V4 | Ideogram | 12 credits per image |
| GPT Image 2.5 | OpenAI | 11 credits per image |
| Flux 2 Max | Black Forest Labs | 8 credits per image |
| Recraft v4 | Recraft | 8 credits per image |
| FLUX.1 Schnell | Black Forest Labs | 1 credits per image |
| Model | Developer | Cost in credits (per second) |
|---|---|---|
| Google Veo 3.1 | 80 credits per second | |
| SeeDance 2.5 | ByteDance | 44 credits per second |
| Kling 3.0 Pro | Kuaishou | 34 credits per second |
| Kling 3.0 Standard | Kuaishou | 25 credits per second |
| Luma Ray 3.2 | Luma | 20 credits per second |
| MiniMax H3 | MiniMax | 10 credits per second |
| Google Veo 3.1 Lite | 10 credits per second | |
| Wan 3.0 | Alibaba | 10 credits per second |
| Model | Developer | Cost in credits (credits) |
|---|---|---|
| ElevenLabs TTS | ElevenLabs | 1 / 10 characters |
| OpenAI TTS | OpenAI | 1 / 100 characters |
| Eleven Music | ElevenLabs | 1 / second |
| Eleven Sound Effects | ElevenLabs | 1 / second |
| Stable Audio 3 (edit & remix) | Stability AI | 9 / track |
| Whisper (transcription) | OpenAI | 1 / second |
| Model | Developer | Cost in credits (per object) |
|---|---|---|
| Meshy 6 | Meshy | 160 credits per object |
| Hunyuan 3D v3.1 Pro | Tencent | 75 credits per object |
| Trellis 2 (image-to-3D) | Microsoft | 60 credits per object |
| Tripo H3.1 | Tripo | 40 credits per object |
| Rodin 2.5 Fast | Hyper3D | 20 credits per object |
| Hunyuan Motion (3D motion) | Tencent | 16 credits per object |
| TripoSR (image-to-3D) | Stability AI & Tripo | 4 credits per object |
Each model has a fixed price in credits, and the same credits work in the app, the API and the MCP. Every account gets 20 free credits every day to try it. Plans start at $9/month for 1,300 credits. Here’s what a small campaign costs:
One subscription instead of separate plans for Midjourney, Runway, ElevenLabs and Freepik.
20 free credits every day
To try it!
$9/
month
For 1,300 credits
Never expire
Unused credits roll over
| Deliverable | Model | Credits |
|---|---|---|
| 10 product images | GPT Image 2.5 | 110 |
| 2 videos, 5 seconds each | MiniMax H3 Max | 160 |
| 30-second music track | Eleven Music | 30 |
| 300-character voiceover | ElevenLabs TTS | 30 |
| Complete mini campaign | 330 |
≈ $2.28 on the Starter plan.
Access 100+ AI models for images, videos, audio, 3D and more. All plans include commercial rights and credits that never expire.
For hobbyists getting started with AI
Billed annually
Get StarterFor content creators and freelancers
Billed annually
Get CreatorFor professionals scaling content production
Billed annually
Get ProCredit-free, always included:
For teams and organizations
Billed annually
Get EnterpriseCredit-free, always included:
The MCP is one way in. The same account, credits and models also work in your own code, your automations and the creative software you already use.
Build Artificial Studio into your own product. One request per generation, with polling or webhooks.
A community node to run any tool inside your automations, or as a tool for an n8n AI Agent.
Generate 3D objects and video without leaving the Blender viewport.
VST3 and AU plugin: generate audio in your session, audition it and drag it onto a track.
Audio Unit plugin: generate sounds and music without leaving Logic.
Use Artificial Studio MCP when:
The Artificial Studio MCP is a remote Model Context Protocol server that connects Claude, ChatGPT and other AI assistants to 100+ AI models for images, video, audio, music and 3D. Once connected, you ask for what you need in the chat and the assistant generates it with your Artificial Studio account.
Copy the MCP URL https://api.artificialstudio.ai/mcp. In Claude.ai or Claude Desktop, go to Customize → Connectors → Add → Add custom connector, paste the URL, click Continue and Add, then sign in to Artificial Studio and approve access. In ChatGPT, turn on Developer mode in Settings → Apps → Advanced settings, create an app with the same URL and choose OAuth. In Claude Code, run: claude mcp add --transport http artificial-studio https://api.artificialstudio.ai/mcp. Cursor, Windsurf and VS Code use the same URL as a remote MCP server. It takes about 2 minutes and needs no API key.
It works with Claude (Claude.ai, Claude Desktop and Claude Code), ChatGPT, Cursor, Windsurf, VS Code and any other MCP client that supports remote HTTP servers.
No. You connect with the URL https://api.artificialstudio.ai/mcp and sign in with your Artificial Studio account the first time. There are no keys to copy or store.
The MCP uses the same credits as the rest of Artificial Studio. Each model has a fixed cost: for example, a GPT Image 2.5 image costs 11 credits and a second of MiniMax H3 Max video costs 16 (768p). Every account gets 20 free credits every day to try it, and plans start at $9/month for 1,300 credits. Credits roll over and never expire.
You can try it for free: every Artificial Studio account gets 20 free credits every day, enough for a couple of images with most models. For regular use, plans start at $9/month for 1,300 credits, and unused credits accumulate from month to month instead of expiring.
100+ models, including Veo 3.1, Kling 3.0, SeeDance 2.5, MiniMax H3, Nano Banana Pro, GPT Image 2.5, Seedream 5.0, Ideogram V4, Flux 2, ElevenLabs, Eleven Music, Meshy 6, Hunyuan 3D and Trellis 2.
Yes. You can pass files that are already online, and in clients that can run commands (like Claude Code) the assistant uploads your local files for you before generating.
Yes. Paid plans include commercial usage rights, so you can deliver what you create to clients, publish it or sell it.
The MCP is for working inside AI assistants and code editors, with a one-line setup and sign-in. The REST API is for building Artificial Studio into your own apps and automations, with an API key. Both use the same account, credits and plan.
Every generation is saved to your Artificial Studio library, so you can find, download or reuse it later from the app or ask your assistant to list recent generations.
Connect once, then ask for images, videos, music, voiceovers and 3D in any conversation.
