I'm building a new next-gen AI companion app
Hey everyone,😎
Solo dev here. I've been working on a new AI companion product, and I wanted to share what I've got so far ,because honestly, I think there's a huge gap in what's currently out there.
Most AI companion apps give you a text box and call it a day. I want to take things a lot further. Here's what I'm building:
Core Chat Modes:
Text Chat — The classic, but done right. Better context handling, richer formatting.
Visual Chat — An animated character on screen that actually moves and reacts — not just a static avatar staring at you. You can also turn on your camera for face-to-face conversations. The AI sees you and responds accordingly.
Voice Chat — Real-time voice conversations with your AI companion.
Multimodal AI Support:
Full multimodal model integration — the AI can recognize images you send and generate/send images back to you.
Third-Party API Integration:
Bring your own API key. Connect to OpenAI, Claude, Gemini, local models, or any other LLM API you want. Zero platform lock-in and you pick the model, you stay in control.
Memory & Identity:
Persistent Memory — Your AI actually remembers past conversations. Not just the last 10 messages — I'm talking long-term, structured memory that carries across sessions.
Dynamic Covers — Character profiles with animated, dynamic covers instead of a boring static image.
Auto Character Card Generation — Pick some tags, write a short prompt, and the system generates a full character card for you. No more spending hours hand-writing JSON.
I know there are plenty of tools out there already sillyTavern, CAI, polybuzz,chai,you name it. But I haven't found a single product that truly nails the full experience: text + visual + voice + memory + multimodal, all under one roof, with a UX that doesn't make you want to pull your hair out.
This is still in active development, which means right now is the best time to help shape what this becomes.
A few questions for you all:
Which of these features matters most to you?
What's the #1 thing missing from your current AI companion setup?
Are there any features you'd love to see that nobody else is doing?
I'll be posting updates as things progress. If any of this sounds interesting, give me a follow so you don't miss out. All feedback is welcome even if it's "this already exists and here's why yours will fail." I'd much rather hear it now than after launch.
Thanks for reading. Let's build something better.