A glass cube turning in hand, generated with Veo 3.1
API

Infrastructure for Visual Intelligence.

Fast, reliable inference from the researchers who understand models from the inside.

45M+Images & videos served
75+Models
13Providers
11Modalities
01Model catalog
VideoKling V3 Pro
VideoVeo 3.1
VideoSora 2 Pro
VideoSeedance 1.5 Pro
VideoMiniMax Hailuo-02
VideoKling 2.6 Pro
VideoGrok Video
VideoKling 2.1 Master
ImageFlux 1.1 Ultra
ImageGPT Image 2 High
ImageFlux 1.1 Pro
ImageSeedream 4.5
AvatarHedra Avatar
AvatarHedra Character 3
AvatarOmnihuman 1.5
AvatarKling AI Avatar v2 Pro

New frontier models land on the API the week they release.

02Quickstart

One API. All the top models.

One workload lifecycle across modalities.

Three example workloads across 75+ production models
cURL · API v3
export HEDRA_API_KEY="<key_id>:<secret>"

curl -X POST "https://api.hedra.com/v3/models/kling-v3" \
  -H "Authorization: Key $HEDRA_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "input": {
    "prompt": "Hands resting beside an espresso at a sunlit Roman cafe",
    "aspect_ratio": "16:9",
    "duration_ms": 5000,
    "quality": "pro"
  }
}'
Hands beside an espresso at a sunlit Roman cafe, generated with Kling V3 Pro
submit202 job
queuescheduled
runprocessing
storeasset_…
observelogs + status

Models are ready-to-run workloads on the same infrastructure for scheduling, execution, storage, logs, and outputs.

Every model behind one integration — routing, scaling, and reliability handled by the stack that serves our own frontier models.

Glass shattering in slow motion, generated with Kling

What will you build?