TokPortal
Article

White-Label Social Media API: Add TikTok & Instagram to Your AI SaaS

Add managed account creation and publishing to an AI SaaS. Compare the API models, test a dry run and design billing, approvals and delivery status.

Vincent Tellenne

Vincent Tellenne

Co-founder & CEO

September 16, 20267 min read
Actual TokPortal interface for configuring video, carousel and story content

TokPortal product capture, September 2026. The approved asset and publishing configuration are separate from creative generation. Options depend on the selected platform and workflow.

Share
On this page
Quick answer

A white-label social media API lets your product sell a branded workflow while another provider supplies part of the infrastructure. Decide whether customers need scheduling for existing accounts or managed account creation and operations. TokPortal supports the latter for TikTok and Instagram; your app still needs billing, approvals, delivery tracking and customer support.

Part of Vincent Tellenne’s agency-building series, adapted from his original white-label TokPortal essay. Worked budgets and sample offers below are illustrations, not customer results.

Your app generates a video. The user downloads it. Then what?

For some customers, an export is enough. For others, the next steps are the hard part: which account to use, who publishes, what gets approved, whether the job finished and how to repeat it next week. If your users keep asking those questions, distribution may belong in your product.

You do not have to build the entire operating layer yourself. You do have to decide which customer problem you are taking responsibility for. Putting an API behind your logo is the beginning of a product decision, not the end of one.

Sources & real examples

Public accounts, original posts and primary references. Each example is linked where it is discussed.

View all 4 sources and review notes

Choose the API model that matches the customer’s accounts

A scheduler API generally helps users connect and publish to accounts they already control. A managed account API can add account creation, operational delivery and publishing work. These solve related but different problems.

If your customers already have established brand accounts and want a calendar, start by evaluating that workflow. If they need new accounts for a specific market or a purpose-built organic content program, evaluate managed operations. Ask this question before designing the onboarding screen; it changes what users must provide and what they expect to receive.

TokPortal’s developer documentation covers managed account and publishing workflows. Compare the jobs, supported endpoints and terms rather than treating every “social media API” as interchangeable.

What white-label means in your product

Your users buy your offer, use your interface and ask your support team for help. The infrastructure provider handles the operations you order. You decide how the workflow fits your product and what the customer-facing plan includes.

That does not automatically give you a ready-made hosted dashboard, a reseller discount or a guarantee of platform reach. Confirm the commercial arrangement and technical scope you need. Your product must accurately describe what it provides and disclose service providers where your agreements and privacy obligations require it.

A useful first feature might be “publish an approved product-demo campaign to managed TikTok accounts.” That is small enough to explain. An unlimited promise to generate, publish and grow every customer on every social platform is much harder to deliver.

Start with a dry run that cannot create a live campaign

TokPortal exposes its REST API under https://app.tokportal.com/api/ext. The sandbox documentation describes the X-TokPortal-Dry-Run: true header for validating write requests before execution. Keep API credentials on your server, never in browser code.

The example below requests a simulated TikTok account-and-video bundle. Supply your key through a secure environment variable. Leave the dry-run header in place while learning the workflow. The response includes validation and pricing information, but simulated identifiers do not represent live resources.

dry-run-example.sh
curl -X POST https://app.tokportal.com/api/ext/bundles \
  -H "X-API-Key: $TOKPORTAL_API_KEY" \
  -H "X-TokPortal-Dry-Run: true" \
  -H "Content-Type: application/json" \
  -d '{
    "bundle_type": "account_and_videos",
    "platform": "tiktok",
    "country": "USA",
    "videos_quantity": 3,
    "edits_quantity": 0,
    "wants_advanced_warming": false
  }'

The current documentation says bundle creation debits credits before publication. That matters for your checkout: a live create action is not a harmless preview. Build an explicit customer confirmation and cost check before allowing your server to perform a paid operation. Use the current response and pricing documentation rather than hardcoding a price from this article.

Design the customer journey before adding bulk buttons

Start with one campaign and one account. The customer chooses the supported platform and market, sees the scoped offer, approves the content and confirms the operation. Your backend records the request and maps the resulting supplier identifiers to the correct customer.

Give the user a useful view of the work: waiting for their input, submitted for processing, needing correction, under review or completed with a public URL when available. Derive those labels from the documented lifecycle. Do not infer that an event containing the word “published” necessarily proves the public post is live.

For TikTok and Instagram, confirm the current bundle options in the API. Treat YouTube separately: the public getting-started documentation describes a team-arranged workflow rather than the same self-serve account-creation path. Keep unavailable choices out of your checkout.

TokPortal product capture, September 2026. The approved asset and publishing configuration are separate from creative generation. Options depend on the selected platform and workflow. Open TokPortal
Actual TokPortal interface for configuring video, carousel and story contentOpen full-resolution image ↗

Build reliable delivery tracking

Use the webhook documentation and event catalog for the supported events and signature scheme. Verify signatures against the exact raw request body and store the event identifier so a retried delivery does not trigger duplicate customer actions.

Map provider objects to your own tenant and order records on the server. A customer should not be able to retrieve another customer’s work by changing an ID in a URL. Keep an audit trail of approval, submission and delivery so support can explain what happened.

Have a reconciliation path when a callback is delayed. Read the documented resource state rather than immediately creating a second order. Keep supplier retries separate from the decision to charge your customer again. Those are different events with different consequences.

For your first release, a small support queue is often more valuable than another generation feature. Show a human which job is waiting, what the customer approved and what evidence is available. Make “needs attention” a real state, not an error buried in a log.

Price recurring operations as recurring costs

An AI SaaS that charges only for generation can lose money when it adds account operations and support. Model setup, ongoing coverage, content production, publishing, payment fees and your own support time separately. A customer with three accounts and daily content creates a different workload from someone who publishes one demo a week.

Prefer explicit allowances in the first plan. Say how many managed accounts, approved assets and publishing jobs are included. Define how overages are quoted and what happens when a payment fails. Avoid “unlimited” until you can explain how the business absorbs heavy usage.

The calculator below can model one customer’s monthly contribution. In a SaaS, replace the labor estimate with observed support and review hours, and allocate the relevant hosting and tooling costs. Customer acquisition, tax and additional shared overhead still need their own model.

Build an offer you can actually deliver

Agency retainer calculator

Editable example in USD. These are invented planning assumptions, not TokPortal prices or a promised agency margin. Include rejected generations in the cost per approved video.

Modeled monthly cost

$1,177.00

Contribution

$623.00

Contribution margin

34.6%

Where the delivery budget goes

  • Production: $540.00
  • Account operations: $160.00
  • Your time: $280.00
  • Other costs: $90.00
  • Contingency: $107.00

Contribution = client price minus production, operations, valued labor, allocated costs and contingency. This is not net profit. Add tax, payment fees, acquisition costs and any overhead not entered above. Quote one-time setup separately or include it when modeling month one.

Keep your offer draft

PILOT OFFER DRAFT — assumptions, not a quote
Customer: Independent hotels with existing room footage
Monthly client price: $1,800.00
Approved videos: 12
Production: $540.00
Account and publishing costs: $160.00
Delivery/support: 8 hours × $35.00
Other allocated costs: $90.00
Contingency: 10% = $107.00
Total modeled monthly cost: $1,177.00
Contribution before tax, sales costs and unmodeled overhead: $623.00
Contribution margin: 34.6%

Before sending: confirm account count, platforms, country, first-month setup costs, usage rights, included revisions, approval owner, payment schedule, reporting, and exit process. Obtain the current supplier quote. No guaranteed reach or sales.

Use AGD as a product use case customers can understand

Account-Generated Distribution (AGD) gives distribution a structure: purpose-built accounts, each with an audience and a recurring content promise. Your SaaS could help an app team turn a library of real product demonstrations into distinct, approved series for those audiences.

For example, a language app might organize travel phrases and workplace phrases as separate editorial programs. The product can guide the user through audience choice, approved claims, asset review and publishing. It should not encourage accounts that falsely pose as independent satisfied customers.

The customer buys a repeatable way to run that program. The API enables it. Make the use case visible in onboarding, templates and documentation so people do not have to invent the workflow after paying.

Copy an integration acceptance checklist

One target customer and campaign use case:
Existing-account scheduling or managed accounts:
Supported platform and country confirmed:
Server-side API secret and tenant isolation:
Dry run validates the order and cost:
Customer approval and paid-action confirmation:
Stable local order mapped to provider identifiers:
Verified webhooks, duplicate-event handling and reconciliation:
Completed delivery shown with evidence when available:
Support and correction queue:
Renewal, overage, exit and handover rules:
No production launch until one end-to-end pilot is verified:

Launch with one design partner and a complete workflow

Choose a customer willing to give detailed feedback on one clearly priced use case. Walk through the full journey with them: offer, asset approval, order, status, delivery, support and renewal. Record where they hesitate and which words confuse them.

A successful API response proves that one request worked. A useful product lets the customer understand the whole job. Before promoting the feature broadly, test an invalid asset, a duplicate callback, a delayed operation and a request for a correction. These are ordinary product situations, not rare edge cases.

Once one customer can use the workflow repeatedly and your cost model holds up, expand the allowance or supported use cases. You will be building from observed demand instead of a speculative list of integrations.

Test the managed distribution API

Validate a request with a dry run, then scope a real customer pilot before enabling paid operations.

Read the sandbox quickstart
Share
Vincent Tellenne

Written by

Vincent Tellenne

Co-founder & CEO

Vincent is a co-founder and CEO of TokPortal. He works on the infrastructure and operating model behind scaled organic social media distribution.

Learn more about this topic with AI