Quickstart & API Overview

Call image generation, multi-image editing, and synced-audio video models through one unified HTTPS JSON API, with usage-based billing ($1 = 100 Credits) and 30-day HTTPS media URLs.

Base URL · https://api.callmodels.dev

Base URLs & 4 core endpoints

Send all API requests to https://api.callmodels.dev with Authorization: Bearer cm_live_... from your server. Generated media files are served from https://files.callmodels.dev via 30-day signed HTTPS URLs.

EndpointPurposeBilling
GET /v1/auth/checkVerify API key status and optional model allowlist access0 Credits (no charge)
POST /v1/images/generationsSubmit text-to-image or 1–10 reference image edit job (HTTP 202)Hold → settle on completion
POST /v1/videos/generationsSubmit text-to-video or first/last-frame video job (HTTP 202)Duration hold → 0.1s settle
GET /v1/tasks/:task_idPoll task status, credit charge & 30-day HTTPS result URL0 Credits (no charge)

Run your first call in 3 steps

1. Create a server-side key (cm_live_...) in Console → API Keys (/console/api-keys) and top up from $10 (1,000 Credits) in Console → Credits (/console/credits).

2. Optionally verify your key at 0 Credits with GET /v1/auth/check?model=seedream-5.5, then submit a generation task with an Idempotency-Key and poll GET /v1/tasks/:id until completed.

# Step 1: Zero-cost key & model check
curl -sS "https://api.callmodels.dev/v1/auth/check?model=seedream-5.5" \
  -H "Authorization: Bearer $CALLMODELS_API_KEY"

# Step 2: Submit async task (HTTP 202 Accepted)
CREATE_JSON=$(curl -sS -X POST "https://api.callmodels.dev/v1/images/generations" \
  -H "Authorization: Bearer $CALLMODELS_API_KEY" \
  -H "Idempotency-Key: qs_demo_20261115_01" \
  -H "Content-Type: application/json" \
  -d '{
    "model": "seedream-5.5",
    "prompt": "Minimalist glass perfume bottle on travertine stone, soft studio rim light",
    "size": "16:9",
    "quality": "1.5K",
    "output_format": "png"
  }')

TASK_ID=$(echo "$CREATE_JSON" | jq -r ".id")

# Step 3: Poll until completed or failed
while true; do
  TASK_JSON=$(curl -sS "https://api.callmodels.dev/v1/tasks/$TASK_ID" \
    -H "Authorization: Bearer $CALLMODELS_API_KEY")
  STATUS=$(echo "$TASK_JSON" | jq -r ".status")
  if [ "$STATUS" = "completed" ] || [ "$STATUS" = "failed" ]; then
    echo "$TASK_JSON" | jq .
    break
  fi
  sleep 2
done

Public model ID directory

Always pass the stable Callmodels model IDs listed below. As platform capacity scales, your request schema, auth header, and response format stay unchanged.

Public Model IDEndpointBilling & Starting Rate
gpt-image-2.6-flarePOST /v1/images/generationsActual tokens · 3.525 Credits / 1K out tok (~1.629 Credits at 16:9 low)
gpt-image-2.6-sunburstPOST /v1/images/generationsActual tokens · 3.525 Credits / 1K out tok (~1.629 Credits at 16:9 low)
seedream-5.5POST /v1/images/generationsFixed per image · 1.88 Credits (1K/1.5K), 3.29 (2K), +0.752 / ref
seedance-2.5-text-to-videoseedance-2.5-image-to-video
POST /v1/videos/generations0.1s duration · 480p 14.061 · 720p 30.239 · 1080p 75.597 Credits/s
seedance-2.5-flash-text-to-videoseedance-2.5-flash-image-to-video
POST /v1/videos/generations0.1s duration · 480p 14.100 · 720p 28.200 Credits/s
seedance-2.0-fast-text-to-videoseedance-2.0-fast-image-to-video
POST /v1/videos/generations0.1s duration · 480p 5.369 · 720p 11.547 Credits/s
seedance-2.0-text-to-videoseedance-2.0-image-to-video
POST /v1/videos/generations0.1s duration · 480p 8.838 · 720p 19.007 · 1080p 47.518 · 4K 96.938 Credits/s
seedance-2.0-mini-text-to-videoseedance-2.0-mini-image-to-video
POST /v1/videos/generations0.1s duration · 480p 2.169 · 720p 4.665 Credits/s