DreamFort
DreamFort is a privacy-focused AI suite that provides unfiltered chat, video, and image generation through a single browser-based or mobile platform. The platform routes all requests through an Oblivious HTTP relay and a TEE-isolated gateway, making it technically impossible to link users to their prompts, while offering access to frontier models like GPT, Claude, Gemini, and Grok alongside video tools like Seedance 2.5 and ReCast for face-swapping into trending clips.
- Artificial Intelligence
- Creator / Creative Economy
- Cybersecurity
- Software Only
Funding
Founders
Founder details are not available yet.
Product
Problem
Mainstream AI chat and generation tools typically log prompts, tie usage to personal identities, and apply restrictive content filters, discouraging users from asking sensitive questions about health, finances, legal matters, or work. This lack of privacy creates a chilling effect and forces users to choose between useful AI assistance and protecting their personal information.
Solution
DreamFort provides a unified AI suite that combines confidential chat, video and image generation, and a character-swap tool called ReCast, all behind an identity-stripping architecture. Every request is encrypted on the user's device, routed through an Oblivious HTTP relay that separates network identity from content, and decrypted only inside a trusted execution environment (TEE) with remote attestation, so no single party can correlate who a user is with what they asked. The platform offers access to frontier chat models including ChatGPT, Claude, Gemini, and Grok, plus abliterated open models like DeepSeek, and a video studio with models such as Seedance 2.5, Kling, MiniMax, and the unrestricted Wan line for generating up to 30-second clips with synced audio. Users can also star in trending memes and movie scenes through ReCast by uploading a single photo, with the platform handling the character swap while preserving the original motion and camera work.
Target Audience
Primary users are privacy-conscious individuals who want unrestricted access to frontier AI models for sensitive personal, professional, or creative questions, as well as content creators seeking unfiltered video and image generation tools without identity tracking.
Features
- Three-layer privacy architecture: on-device encryption, OHTTP relay (RFC 9458), and TEE-isolated gateway with remote attestation, making prompt-to-identity correlation technically impossible
- Access to 50+ models in one app, including GPT, Claude, Gemini, Grok, DeepSeek, Seedance 2.5, Kling v3, MiniMax H3, Wan, Nano Banana 2, GPT Image 2, and Seedream
- Video Studio with AI director agent that plans scenes, writes per-shot prompts, and chains clips into multi-scene films up to 30 seconds with synced audio and flexible aspect ratios
- ReCast character-swap tool supporting up to four roles per scene, with optional character sheets for full outfit changes and clips up to one minute stitched over the original soundtrack
- Local-only chat history stored in the browser or mobile app, with no remote conversation storage on DreamFort servers
- Cross-platform availability via web browser, Android app on Google Play, and iOS app on the App Store, with voice-to-text and text-to-speech capabilities