Skip to main content

API Services Overview

한국어로 보기: API 서비스 개요 | View in English (current page)

kvidAI exposes a unified HTTPS API at https://api.kvid.ai. Every service shares a single credit balance and a single API key issued from the web app.

Services at a Glance​

ServiceDescriptionReference
Video GenerationText-to-video, image-to-video, and reference-to-video (wan / seedance / veo3.1 models)Video API
Image GenerationText-to-image and image-to-image (nano-banana / flux / sdxl)Image API
Voice (TTS)ElevenLabs-based text-to-speech with character-level timing for subtitle syncVoice API
Speech-to-TextElevenLabs Scribe transcription (words + timestamps) from a file or CDN URLSpeech-to-Text API
AI EditMedia summary (STT + LLM) and silence-cut, streamed over SSEAI Edit API
Agent (AI Editor)Natural-language composition editing and long-video scene planning via SSEAgent API
Project ManagementREST CRUD for video projects + composition mutationsProject Management API
PresetReusable presets (voice / tone / color / scene defaults) that seed new projectsPreset API
MediaDirect CDN upload via presigned URL — pass cdnUrl to the agent without round-tripping the binary through the serverMedia API

Getting Started​

1. Sign up and buy credits​

Everything happens in one place:

  1. Create an account at kvid.ai (email verification required).
  2. Buy credits at kvid.ai/credits/purchase.

2. Create an API key​

Go to kvid.ai/settings/api-keys and create a new API subscription. You will receive a primary and secondary key; either can be used with the API.

3. Make your first call​

curl -X POST "https://api.kvid.ai/ai/generation/text-to-image/generate-async" \
  -H "api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "product_id": "YOUR_PRODUCT_ID",
    "prompt": "K-pop concert stage with colorful lights",
    "model": "nano-banana"
  }'

The call returns a job_id. Poll GET /ai/generation/status?jobId={id} until status: "completed", then fetch the result with GET /ai/generation/result?jobId={id}. See the Image API reference for details.

Identifying the credit pool — the AI generation endpoints (image / video / talk-v2v / voice) require one of product_id, product_code, or email in the request body so the platform knows which credit balance to charge. The platform endpoints (Preset, Media, Project, Agent, AI Edit) don't need this — the api-key header alone identifies you.

Pricing​

All services are priced on the shared credit balance. See Pricing for the current rates.

K-pop & K-beauty Specialization​

  • Video: optimized prompting for idol choreography, stage camera angles, and Korean cultural contexts
  • Image: K-beauty makeup and Korean fashion styling baked into model tuning

Support​