API reference

Three steps from zero to a finished file. You need a workspace (created with your account) and an API key with the scopes for the endpoints you call.

1. 1

   Create a scoped key

   In Dashboard → API keys, pick only the scopes you need. The plaintext key is shown once.

2. 2

   Read the model catalog

   Call GET /v1/models with your workspaceId. Copy the model id and note its parameterSchema.

3. 3

   Submit, then poll or subscribe

   POST /v1/generations returns 201 immediately. Poll the detail endpoint or wait for a signed webhook.

export FREEIND_API_KEY="fi_live_..."
export WORKSPACE_ID="00000000-0000-4000-8000-000000000000"
export BASE="https://free.ind.in"

# 1 — pick a model and copy its id
curl -s "$BASE/v1/models?workspaceId=$WORKSPACE_ID" \
  -H "Authorization: Bearer $FREEIND_API_KEY"

# 2 — submit a generation (201 = queued, not finished)
curl -s -X POST "$BASE/v1/generations" \
  -H "Authorization: Bearer $FREEIND_API_KEY" \
  -H "Idempotency-Key: hero-image-001" \
  -H "Content-Type: application/json" \
  -d '{
    "workspaceId": "'"$WORKSPACE_ID"'",
    "modelId": "<model-id>",
    "params": { "mode": "text_to_image", "prompt": "a red fox in fresh snow" }
  }'

# 3 — poll until succeeded, or wait for the generation.completed webhook
curl -s "$BASE/v1/generations/<generation-id>/raw?workspaceId=$WORKSPACE_ID" \
  -H "Authorization: Bearer $FREEIND_API_KEY"

Every free.* tool is available over the same API. Generation-based tools (reel, edit, fill, animate, thumbnail, try-on) use POST /v1/generations with the matching mode; the pipeline tools below have dedicated endpoints.

Pipeline tools need the generations.* scopes; free.flow and reference templates also need workflows.read / workflows.write. Chat accepts any valid key.

Every request authenticates with a scoped API key created in the dashboard. Send it as a bearer token — never in a query string, never in client-side code.

curl "https://free.ind.in/v1/models?workspaceId=$WORKSPACE_ID" \
  -H "Authorization: Bearer $FREEIND_API_KEY"

Keys are bound to one workspace. Requests whose workspaceId belongs to a different workspace fail with 403 forbidden.

Session endpoints (API keys, billing, workspaces, account) authenticate with the dashboard cookie instead and are marked as such below.

Failures always return a JSON envelope with a stable machine-readable code and the requestId to quote in support requests.

{
  "error": {
    "code": "validation_error",
    "message": "Invalid request body",
    "details": [{ "field": "params.prompt", "message": "Required" }],
    "requestId": "req_01J9Z8K2"
  }
}

List endpoints accept page (1-indexed) and perPage, and return an items array with pagination metadata:

{
  "items": [{ "id": "…" }],
  "meta": { "hasNextPage": true, "nextCursor": null, "total": null }
}

Keep requesting until hasNextPage is false. Cursor pagination (cursor, nextCursor) is reserved; the credit ledger uses page + limit.

Limits are enforced per API key and come from the workspace plan entitlements (default 60 requests/minute and 1,000/day). Every API-key response carries the current window:

X-RateLimit-Limit: 60
X-RateLimit-Remaining: 57
X-RateLimit-Reset: 42

A 429 rate_limited response means the window is exhausted. Back off until X-RateLimit-Reset seconds have passed instead of retrying immediately.

Send Idempotency-Key (8–128 chars of [A-Za-z0-9._:-]) on POST requests so a retry after a timeout cannot create a duplicate job. Keys are scoped to the workspace; reusing one returns 409 idempotency_conflict instead of charging twice.

curl -X POST "https://free.ind.in/v1/generations" \
  -H "Authorization: Bearer $FREEIND_API_KEY" \
  -H "Idempotency-Key: order-1042-scene-3" \
  -H "Content-Type: application/json" \
  -d @generation.json

Each delivery is a POST with an X-FreeInd-Signature header shaped t=<unix>,v1=<hex>. Verify HMAC-SHA256 over t + "." + rawBody using the secret returned at creation, then reject anything older than five minutes.

import { createHmac, timingSafeEqual } from 'node:crypto';

// header: "t=1761307218,v1=8f2c..."
export function verifyFreeIndSignature(secret, rawBody, header) {
  const [timestampPart, signaturePart] = header.split(',');
  const timestamp = timestampPart.slice(2);
  const signature = signaturePart.slice(3);

  const expected = createHmac('sha256', secret)
    .update(`${timestamp}.${rawBody}`)
    .digest('hex');

  const valid = timingSafeEqual(Buffer.from(signature), Buffer.from(expected));
  const fresh = Math.abs(Date.now() / 1000 - Number(timestamp)) < 300;
  return valid && fresh;
}

Event types: generation.completed, generation.failed, generation.cancelled, generation.expired, workflow.completed, workflow.failed, workflow.cancelled. Deliveries retry with backoff; inspect them with the deliveries endpoint.

Import the collection to get every API-key endpoint with folders, variables, request examples, and response tests that chain a created generation id into the follow-up requests.

Import in Postman:

1. Postman → Import → Link and paste https://free.ind.in/downloads/freeind-api.postman_collection.json (or drop the downloaded file).
2. Select the “FreeInd API — Production” environment.
3. Set apiKey, workspaceId, and modelId.

A hand-curated OpenAPI 3.1 document powers this page. Point Swagger UI, Insomnia, Stoplight, or an LLM at it directly:

• https://free.ind.in/openapi.json — curated spec with descriptions, request bodies, and response examples.
• https://free.ind.in/api/docs/json — the raw spec emitted by the live Fastify server (route inventory, no examples).

The API is versioned under /v1. Breaking changes ship with a 90-day deprecation window — see the changelog policy.