I need a developer to build a clean, responsive web application that lets me interact with three AI models (Claude, ChatGPT, and Gemini) from a single interface. Important context: This is not a commercial product, not a SaaS, and not consumer-facing. This is a personal automation tool for myself only. One user, no accounts, no scaling. Please scope and price your work accordingly — I don't need enterprise architecture, user management, or anything over-engineered. I just need it to work well and be easy to maintain. The Problem I'm Solving: I use Claude, ChatGPT, and Gemini daily. I manually type or paste the same prompt into all three platforms, then I copy each AI's response into another AI to get them to react to each other's thinking. This is exhausting, especially on mobile. I need an app that eliminates this entirely. There are two core functionalities: FEATURE 1: Unified Prompt, Separate Conversations A single input box that sends my message to all three AI APIs (Anthropic Claude, OpenAI ChatGPT, Google Gemini) at the same time. Each model has its own separate response panel three distinct columns on desktop, three stacked panels on mobile. Each model's conversation history is displayed independently in its own panel, like three separate chat windows that just happen to receive the same input from me. This is NOT a single-prompt tool. This is a full, persistent conversation interface. Every message I type goes to all three models simultaneously and each model maintains its own independent conversation context throughout. To be very clear about the expected behavior: I open a new conversation and type my first message → it goes to Claude, ChatGPT, and Gemini simultaneously → I see three separate responses, each in its own panel I type a follow-up → that follow-up goes to all three, each with full memory of its own conversation so far → I see three updated responses in their respective panels This continues for as many turns as I want — the fifth message, the tenth message, every message I type always goes to all three Each model's panel shows its own complete conversation thread, like three independent chat windows side by side I can start a fresh conversation at any time, which resets context for all three models Conversation history is saved so I can go back and review past conversations FEATURE 2: Cross-Feed Responses Between Models At any point after all three models have responded, I can trigger a "cross-feed" action. This takes each model's latest response and sends it to the other two models simultaneously so they can react to each other's thinking. Here is exactly what happens when I trigger this: Claude receives a message containing ChatGPT's latest response and Gemini's latest response → Claude processes and responds ChatGPT receives a message containing Claude's latest response and Gemini's latest response → ChatGPT processes and responds Gemini receives a message containing Claude's latest response and ChatGPT's latest response → Gemini processes and responds All three of these fire at the same time. No waiting for one to finish before sending to the next. Each AI gets the other two's latest outputs and all three generate their updated responses simultaneously. This should be a single button or action one tap and it happens. I should be able to trigger this multiple rounds in a row if I want the models to keep refining based on each other's latest thinking. Each round of cross-feeding becomes part of that model's ongoing conversation context. FEATURE 3: Ability to export any conversation (or all conversations) as a downloadable file. This is just a copy — exporting does not delete anything from the app. I want to be able to save conversations externally for my own records or import them into another tool later if needed. A simple file format like JSON or markdown is fine. UI/UX Requirements: Clean, minimal interface. Power-user tool, not a consumer product. Fully responsive must work smoothly on mobile browsers (iPhone and Android). I use this heavily on my phone. Three separate panels for the three models columns on desktop, stacked on mobile. Each panel shows that model's full conversation thread independently. Real-time streaming so I see responses as they generate, not after they finish Dark mode preferred One-tap copy for any individual response Conversation list/sidebar to switch between past and new conversations Clear visual distinction between my messages, each model's responses, and cross-fed rounds Model Selection: I need to be able to choose which specific model from each provider (e.g., Claude Opus vs Sonnet, GPT-4o vs o1, Gemini Pro vs Flash) API keys stored locally (single-user tool, no need for a backend auth system) Technical Requirements: Web app (hosted on Vercel, Cloudflare Pages, or similar I'll provide hosting) Modern frontend framework (React/Next.js preferred) No backend database local storage ONLY Clean, well-commented code so it can be maintained and extended later Graceful handling of API rate limits and errors (retry logic, clear error messages) Ability to see how much tokens or money I've spent generally About Me: I need this tool to work reliably. I don't need anything fancy I need it fast, stable, and frictionless. I will provide all API keys. Again, this is just a personal automation tool for myself, so please keep your approach simple and practical. To Apply, Please Include: Your experience working with LLM APIs (specifically Anthropic, OpenAI, and/or Google Gemini) A link to any relevant past project — a chatbot, AI tool, or similar web app Your estimated timeline and cost Your preferred tech stack and why