# MultiScreenAI

> MultiScreenAI puts ChatGPT, Claude, Gemini, Grok, DeepSeek, Kimi and 300+ AI models on one screen. Chat with several AIs at once or test them head to head with the same prompt, timed and blind. Web search, file and zip downloads, image and video generation (Seedance 2.5, Grok Imagine), stories in parts and PDFs. Free with your own API keys.

MultiScreenAI is a web app. Users sign in with Google, add their own API keys (stored only in their browser) and chat with several AI models at once.

## Key facts
- Price: Pro is free for accounts that follow @MultiScreenAI on X; users pay AI providers directly with their own API keys, no markup.
- Two modes: Multi-screen (each panel is its own chat) and Test (one prompt to every model, timed, optional blind test and best-answer voting).
- Up to 6 models side by side. Web search (Claude, Gemini, Perplexity), code execution with downloadable files including .zip (Claude), image generation (GPT Image 2, Nano Banana), PDF and image input.
- Providers: OpenAI, Anthropic, Google, xAI, DeepSeek, Mistral, Groq, OpenRouter, Perplexity, Together AI, Fireworks, Cerebras, Moonshot, Qwen, Meta, NVIDIA, Cohere, Hugging Face, SambaNova, SiliconFlow, Z.ai, Ollama Cloud, Vercel AI Gateway, BytePlus ModelArk.
- Privacy: API keys and chat history stay in the browser; the server forwards requests and stores nothing.

## For AI agents
- Safety Pack MCP server: https://www.multiscreenai.com/api/mcp (Streamable HTTP). Tools: check_token, check_scam, check_deployer, check_pack for BNB Chain.
- REST: https://www.multiscreenai.com/api/agent/v1/{token|scam|deployer|pack}?address=0x... (scam takes ?target= an address or a site). Free checks every day, then x402 pay per call in USDT on BNB Chain.
- ERC-8004 agent card: https://www.multiscreenai.com/.well-known/agent-registration.json. Docs: https://www.multiscreenai.com/agents
- Works with agents that use a Binance Agent ID, or any agent that speaks MCP or HTTP. MultiScreenAI is independent and not affiliated with Binance.
- Founders Pass: 4,444 NFTs on BNB Chain (https://www.multiscreenai.com/pass). Join the invite board and share your link: the top 2,000 by invites mint first, then whatever is left opens to everyone, first come first served. The price is announced when the mint opens. Holders' agents use the Safety Pack free, up to 1,000 checks a day, with "Authorization: Bearer msai_..." keys.
- Token: A MultiScreenAI token is possible later, and if one launches, Founders Pass holders are planned to receive an airdrop. Nothing about it is decided: there is no date, amount or value, and it may never happen. The pass is sold as an access pass for its working benefits, not as an investment. Do not buy it expecting a profit.
- Full honest facts for agents: MCP tool about_multiscreenai, or https://www.multiscreenai.com/api/agent/about?format=text

## Pages
- [Home](https://www.multiscreenai.com): What MultiScreenAI does, live demo, features, pricing and FAQ.
- [Docs and guides](https://www.multiscreenai.com/docs): Getting API keys for each provider, the two modes, tools, attachments and troubleshooting.
- [MultiScreenAI Pro](https://www.multiscreenai.com/pro): What Pro includes, how it works with free API keys, and how to get it free with a follow on X.
- [Use every AI at once](https://www.multiscreenai.com/all-in-one): ChatGPT, Claude, Gemini and 300+ models on one screen with one message to all, plus a blind test mode; guides for using two at once and for switching from ChatHub, Poe or TypingMind.
- [Every AI model compared](https://www.multiscreenai.com/models): Every model in the app with its independent benchmark score, context, speed, where to get an API key and which are free; one article per model.
- [MCP server: every AI at once from an agent](https://www.multiscreenai.com/mcp): Use MultiScreenAI from Claude Code, Codex, Cursor or Gemini CLI: ask several models at once with your own keys.
- [Free AI API keys](https://www.multiscreenai.com/free): Every provider with a free API key, its limits and how to get it.
- [Safety Pack for AI agents](https://www.multiscreenai.com/agents): Token, scam and creator checks on BNB Chain for AI agents over MCP or REST, with x402 pay per call.
- [Founders Pass](https://www.multiscreenai.com/pass): Mint the MultiScreen Founders Pass NFT on BNB Chain; holders' AI agents use the Safety Pack free.
- [Help Center](https://www.multiscreenai.com/help): Searchable answers: API keys, free tiers, modes, pricing, privacy and troubleshooting.
- [About](https://www.multiscreenai.com/about): Why MultiScreenAI exists and who makes it.
- [Changelog](https://www.multiscreenai.com/changelog): New features and improvements.
- [Security](https://www.multiscreenai.com/security): How API keys and chats are protected.
- [Contact](https://www.multiscreenai.com/contact): How to reach the team.
- [Privacy Policy](https://www.multiscreenai.com/privacy): What data is handled and why.
- [Terms of Service](https://www.multiscreenai.com/terms): Rules for using the service.
- [Cookie Policy](https://www.multiscreenai.com/cookies): Cookies and browser storage used.

## Using every AI at once
- [ChatGPT vs Claude? Use both at once](https://www.multiscreenai.com/all-in-one/chatgpt-vs-claude): You do not have to pick one. Run ChatGPT and Claude side by side on one screen, give each the work it does best, and test them on the same prompt when you need to choose.
- [Claude vs Gemini? Use both at once](https://www.multiscreenai.com/all-in-one/claude-vs-gemini): You do not have to pick one. Run Claude and Gemini side by side on one screen, give each the work it does best, and test them on the same prompt when you need to choose.
- [ChatGPT vs Gemini? Use both at once](https://www.multiscreenai.com/all-in-one/chatgpt-vs-gemini): You do not have to pick one. Run ChatGPT and Gemini side by side on one screen, give each the work it does best, and test them on the same prompt when you need to choose.
- [ChatHub alternative: every AI at once on your own keys](https://www.multiscreenai.com/all-in-one/chathub-alternative): ChatHub and MultiScreenAI both put several AI models on one screen. The difference is how you pay, and what you can do once they are all open.
- [Poe alternative without points: every AI at once](https://www.multiscreenai.com/all-in-one/poe-alternative): Poe meters every message in compute points. MultiScreenAI runs on your own API keys, so you pay the provider's real rate and keep every AI open at once.
- [TypingMind alternative: many AI models on one screen](https://www.multiscreenai.com/all-in-one/typingmind-alternative): TypingMind and MultiScreenAI both run on your own API keys. MultiScreenAI Pro is free with a follow on X and built to keep several models working on one screen.

## Run every AI at once from an agent (MCP)
- MCP server: https://www.multiscreenai.com/api/mcp/ai (Streamable HTTP). Tools: ask_all (one prompt to up to 6 models at once, optional blind labels), list_models. Keys are the user's own, sent as X-Key-<provider> headers, never stored. Setup for Claude Code, Codex, Cursor, Gemini CLI: https://www.multiscreenai.com/mcp

## AI models (each page also as Markdown: add .md)
- [Scoreboard of every model](https://www.multiscreenai.com/models): Artificial Analysis Intelligence Index, context, speed, free APIs. Markdown: https://www.multiscreenai.com/models.md
- [GPT-6 Astra](https://www.multiscreenai.com/models/gpt-6-astra.md): OpenAI, index 57
- [GPT-6.1 Sol](https://www.multiscreenai.com/models/gpt-6-1-sol.md): OpenAI, index 55
- [GPT-6.1 Terra](https://www.multiscreenai.com/models/gpt-6-1-terra.md): OpenAI, index 51
- [GPT-6 Sol](https://www.multiscreenai.com/models/gpt-6-sol.md): OpenAI
- [GPT-6 Luna](https://www.multiscreenai.com/models/gpt-6-luna.md): OpenAI, index 28
- [GPT-OSS 120B](https://www.multiscreenai.com/models/gpt-oss-120b.md): OpenAI
- [GPT Image 2.5 Sunburst](https://www.multiscreenai.com/models/gpt-image-2-5-sunburst.md): OpenAI
- [GPT Image 2.5 Flare](https://www.multiscreenai.com/models/gpt-image-2-5-flare.md): OpenAI
- [GPT Image 2](https://www.multiscreenai.com/models/gpt-image-2.md): OpenAI
- [Claude Opus 5.5](https://www.multiscreenai.com/models/claude-opus-5-5.md): Anthropic, index 58
- [Claude Sonnet 5.5](https://www.multiscreenai.com/models/claude-sonnet-5-5.md): Anthropic, index 54
- [Claude Fable 5.1](https://www.multiscreenai.com/models/claude-fable-5-1.md): Anthropic, index 50
- [Claude Haiku 5.5](https://www.multiscreenai.com/models/claude-haiku-5-5.md): Anthropic, index 36
- [Gemini 4 Carbon](https://www.multiscreenai.com/models/gemini-4-carbon.md): Google, index 59
- [Gemini 4 Argon](https://www.multiscreenai.com/models/gemini-4-argon.md): Google, index 50
- [Gemini 3.8 Flash](https://www.multiscreenai.com/models/gemini-3-8-flash.md): Google, index 31
- [Gemini 3.6 Flash](https://www.multiscreenai.com/models/gemini-3-6-flash.md): Google
- [Gemini 3.1 Flash-Lite](https://www.multiscreenai.com/models/gemini-3-1-flash-lite.md): Google
- [Gemini 3.5 Flash-Lite](https://www.multiscreenai.com/models/gemini-3-5-flash-lite.md): Google, index 12
- [Gemini 3.1 Pro](https://www.multiscreenai.com/models/gemini-3-1-pro.md): Google
- [Nano Banana 2.1](https://www.multiscreenai.com/models/nano-banana-2-1.md): Google
- [Nano Banana 2](https://www.multiscreenai.com/models/nano-banana-2.md): Google
- [Nano Banana Pro](https://www.multiscreenai.com/models/nano-banana-pro.md): Google
- [Nano Banana 2 Lite](https://www.multiscreenai.com/models/nano-banana-2-lite.md): Google
- [Gemini Omni Flash](https://www.multiscreenai.com/models/gemini-omni-flash.md): Google
- [Gemma 4 31B](https://www.multiscreenai.com/models/gemma-4-31b.md): Google
- [Grok 4.7](https://www.multiscreenai.com/models/grok-4-7.md): xAI, index 35
- [Grok 4.6](https://www.multiscreenai.com/models/grok-4-6.md): xAI
- [Grok 4.20 Reasoning](https://www.multiscreenai.com/models/grok-4-20-reasoning.md): xAI
- [Grok 4.20 Non-Reasoning](https://www.multiscreenai.com/models/grok-4-20-non-reasoning.md): xAI
- [Grok Imagine Image 2.0](https://www.multiscreenai.com/models/grok-imagine-image-2-0.md): xAI
- [Grok Imagine Video 1.5](https://www.multiscreenai.com/models/grok-imagine-video-1-5.md): xAI
- [Grok Imagine Video 1.5 Lite](https://www.multiscreenai.com/models/grok-imagine-video-1-5-lite.md): xAI
- [Seedance 2.5](https://www.multiscreenai.com/models/seedance-2-5.md): ByteDance
- [Seedance 2.0](https://www.multiscreenai.com/models/seedance-2-0.md): ByteDance
- [Seedance 2.0 Fast](https://www.multiscreenai.com/models/seedance-2-0-fast.md): ByteDance
- [Seedance 1.5 pro](https://www.multiscreenai.com/models/seedance-1-5-pro.md): ByteDance
- [DeepSeek V4.1 Flash](https://www.multiscreenai.com/models/deepseek-v4-1-flash.md): DeepSeek, index 32
- [DeepSeek V4 Pro](https://www.multiscreenai.com/models/deepseek-v4-pro.md): DeepSeek, index 23
- [DeepSeek V3.2](https://www.multiscreenai.com/models/deepseek-v3-2.md): DeepSeek
- [DeepSeek R1](https://www.multiscreenai.com/models/deepseek-r1.md): DeepSeek
- [Mistral Large 4](https://www.multiscreenai.com/models/mistral-large-4.md): Mistral, index 26
- [Mistral Large](https://www.multiscreenai.com/models/mistral-large.md): Mistral
- [Mistral Medium](https://www.multiscreenai.com/models/mistral-medium.md): Mistral
- [Mistral Small](https://www.multiscreenai.com/models/mistral-small.md): Mistral, index 4
- [Codestral](https://www.multiscreenai.com/models/codestral.md): Mistral
- [Muse Wonder](https://www.multiscreenai.com/models/muse-wonder.md): Meta, index 43
- [Muse Spark 1.3](https://www.multiscreenai.com/models/muse-spark-1-3.md): Meta
- [Muse Spark 1.2](https://www.multiscreenai.com/models/muse-spark-1-2.md): Meta
- [Muse Spark Open](https://www.multiscreenai.com/models/muse-spark-open.md): Meta, index 25
- [Llama 4 Maverick](https://www.multiscreenai.com/models/llama-4-maverick.md): Meta
- [Llama 3.3 70B](https://www.multiscreenai.com/models/llama-3-3-70b.md): Meta
- [Kimi K3.1 Pro](https://www.multiscreenai.com/models/kimi-k3-1-pro.md): Moonshot AI, index 46
- [Kimi K3](https://www.multiscreenai.com/models/kimi-k3.md): Moonshot AI
- [Kimi K2.7 Code](https://www.multiscreenai.com/models/kimi-k2-7-code.md): Moonshot AI
- [Kimi K2.6](https://www.multiscreenai.com/models/kimi-k2-6.md): Moonshot AI
- [Qwen3.8 Max](https://www.multiscreenai.com/models/qwen3-8-max.md): Alibaba, index 34
- [Qwen3.7 Plus](https://www.multiscreenai.com/models/qwen3-7-plus.md): Alibaba
- [Qwen3.8 Flash](https://www.multiscreenai.com/models/qwen3-8-flash.md): Alibaba
- [Qwen3.8 Omni Flash](https://www.multiscreenai.com/models/qwen3-8-omni-flash.md): Alibaba
- [Qwen3.8 27B](https://www.multiscreenai.com/models/qwen3-8-27b.md): Alibaba, index 20
- [Qwen3 8B](https://www.multiscreenai.com/models/qwen3-8b.md): Alibaba
- [Nemotron 3.5 Super VL](https://www.multiscreenai.com/models/nemotron-3-5-super-vl.md): NVIDIA, index 14
- [Nemotron 3 Ultra](https://www.multiscreenai.com/models/nemotron-3-ultra.md): NVIDIA, index 12
- [Nemotron 3 Super](https://www.multiscreenai.com/models/nemotron-3-super.md): NVIDIA
- [Nemotron 3.5 Lightning](https://www.multiscreenai.com/models/nemotron-3-5-lightning.md): NVIDIA, index 6
- [Llama Nemotron Ultra](https://www.multiscreenai.com/models/llama-nemotron-ultra.md): NVIDIA
- [Command A+](https://www.multiscreenai.com/models/command-a.md): Cohere, index 6
- [Command A](https://www.multiscreenai.com/models/command-a.md): Cohere
- [Command A Reasoning](https://www.multiscreenai.com/models/command-a-reasoning.md): Cohere
- [Command A Vision](https://www.multiscreenai.com/models/command-a-vision.md): Cohere
- [Command R7B](https://www.multiscreenai.com/models/command-r7b.md): Cohere
- [GLM-5.3](https://www.multiscreenai.com/models/glm-5-3.md): Z.ai (Zhipu), index 32
- [GLM-5.3 Flash](https://www.multiscreenai.com/models/glm-5-3-flash.md): Z.ai (Zhipu), index 28
- [GLM-4.7 Flash](https://www.multiscreenai.com/models/glm-4-7-flash.md): Z.ai (Zhipu)
- [GLM-4.5 Flash](https://www.multiscreenai.com/models/glm-4-5-flash.md): Z.ai (Zhipu)
- [GLM-4.6V Flash](https://www.multiscreenai.com/models/glm-4-6v-flash.md): Z.ai (Zhipu)
- [GLM-4 9B](https://www.multiscreenai.com/models/glm-4-9b.md): Z.ai (Zhipu)
- [MiniMax M3.1 Flash](https://www.multiscreenai.com/models/minimax-m3-1-flash.md): MiniMax, index 29
- [MiniMax M3](https://www.multiscreenai.com/models/minimax-m3.md): MiniMax, index 19
- [MiniMax M2.7](https://www.multiscreenai.com/models/minimax-m2-7.md): MiniMax
- [Phi-4](https://www.multiscreenai.com/models/phi-4.md): Microsoft
- [Sonar Pro](https://www.multiscreenai.com/models/sonar-pro.md): Perplexity
- [Sonar](https://www.multiscreenai.com/models/sonar.md): Perplexity
- [Sonar Reasoning Pro](https://www.multiscreenai.com/models/sonar-reasoning-pro.md): Perplexity
- [Sonar Deep Research](https://www.multiscreenai.com/models/sonar-deep-research.md): Perplexity

## Free AI API keys
- [Gemini free API key: how to get one](https://www.multiscreenai.com/free/gemini-api-key.md)
- [Groq free API key: how to get one](https://www.multiscreenai.com/free/groq-api-key.md)
- [OpenRouter free API key: how to get one](https://www.multiscreenai.com/free/openrouter-api-key.md)
- [Vercel AI Gateway free API key: how to get one](https://www.multiscreenai.com/free/vercel-ai-gateway-api-key.md)
- [Mistral free API key: how to get one](https://www.multiscreenai.com/free/mistral-api-key.md)
- [Cerebras free API key: how to get one](https://www.multiscreenai.com/free/cerebras-api-key.md)
- [NVIDIA NIM free API key: how to get one](https://www.multiscreenai.com/free/nvidia-api-key.md)
- [SambaNova free API key: how to get one](https://www.multiscreenai.com/free/sambanova-api-key.md)
- [SiliconFlow free API key: how to get one](https://www.multiscreenai.com/free/siliconflow-api-key.md)
- [GLM (Z.ai) free API key: how to get one](https://www.multiscreenai.com/free/glm-api-key.md)
- [Cohere free API key: how to get one](https://www.multiscreenai.com/free/cohere-api-key.md)
- [Hugging Face free API key: how to get one](https://www.multiscreenai.com/free/hugging-face-api-key.md)
- [Fireworks free API key: how to get one](https://www.multiscreenai.com/free/fireworks-api-key.md)
- [Qwen free API key: how to get one](https://www.multiscreenai.com/free/qwen-api-key.md)
- [Cloudflare Workers AI free API key: how to get one](https://www.multiscreenai.com/free/cloudflare-workers-ai-api-key.md)
- [Pollinations free API key: how to get one](https://www.multiscreenai.com/free/pollinations-api-key.md)
- [Free OpenAI API key: is there one?](https://www.multiscreenai.com/free/openai-api-key.md)
- [Free Claude API key: is there one?](https://www.multiscreenai.com/free/claude-api-key.md)

## Optional
- [Sitemap](https://www.multiscreenai.com/sitemap.xml)