Fliki?

NOT REALLYA prototype of the interface is possible, but the durable product depends on difficult systems beyond the UI. Build a safer adjacent workflow instead.

BUILD FIRST

A non-sensitive or read-only version of the core experience
A manual review step wherever correctness or trust matters
A narrow companion tool that removes a real piece of work

LEAVE FOR LATER

Regulated decisions, payments, or irreversible actions
Claims of correctness, compliance, or uptime that the MVP cannot support
Sensitive integrations until security, support, and auditability are ready

RESEARCH NOTES

What this tool actually does.

Checked 2026-09-06

Fliki is a browser-based AI video creation suite. Users paste a script, blog post, or one-line prompt; the system auto-writes a script, pairs each scene with stock or AI-generated footage, synthesizes voiceover from 2,000+ voices in 80+ languages, burns captions in 100+ languages, and adds auto-ducked royalty-free music. It aggregates models like Veo, Sora, Kling, and Seedance under one interface with a timeline editor for trimming, regenerating scenes, and layering text/stickers. Also included: voice cloning, AI avatars, photo-to-video, video translation, one-click resizing to 9:16, 1:1, 16:9, brand kits, templates, and team workspaces with version history. Exports MP4 up to 4K, with direct publishing to YouTube/TikTok/Reels/Shorts/LinkedIn. Free tier offers monthly credits with watermark; paid plans remove it.

SOURCES

RESEARCHED BUILD PROMPT

Give this to your coding agent.

Build a functional replacement for Fliki's core text-to-video workflow, with original branding, UI, and implementation. Workflow: user pastes a script (or a one-line idea auto-expands into a full script via an LLM call); the app splits the script into scenes, auto-suggests a stock clip or generated image per scene, synthesizes voiceover per scene, burns word-level captions, adds background music ducked under narration, and exports a single MP4. User-facing requirements: script editor with server-side AI expansion; per-scene visual picker using a curated royalty-free clip/image subset plus an optional generate-on-demand image endpoint; voice picker across 10+ languages with speed and pitch controls; caption styling (font, size, color, position, animation toggle); music picker with auto-duck toggle; aspect-ratio presets 16:9, 1:1, 9:16 with re-render; preview player; export job queue with visible progress and a mock premium tier that removes the watermark. Technical requirements: React frontend with a Node or Python API; SQLite (or Postgres) storing project scenes, choices, and exports; background worker process for export jobs; TTS via an OpenAI-compatible endpoint or local fallback; video assembly with ffmpeg/libx264 including caption burn-in via drawtext or ASS subtitles; media source as a static royalty-free library plus user uploads; progress events over WebSocket or SSE. Implementation approach: server-side scene pipeline (script → scenes → visual → audio → caption burn → mix) rather than client-side rendering. Explicitly out of scope: multi-user workspaces, voice cloning, AI avatars, meta video model aggregation (Sora/Veo/Kling), social media publishing, brand kits, and translation of existing videos. Do not copy Fliki logos, screens, colors, copy, or layout; this must read as an independent tool. Include a README covering setup, TTS/storage configuration, and important limitations: LLM token limits on long scripts, watermarking and render-time tradeoffs, background-music ducking requires sidechain or amplitude analysis, and caption burn-in font licensing.

Visit the official product page