Speechify API Pricing Explained: App, Studio, and API

Short answer: Do not assume Speechify's Reader app, Speechify Studio or voiceover workspace, and SpeechifyAI API share one billing model or allowance. Speechify explicitly describes Reader and Studio as two different subscriptions, lists the API as another product, and publishes a separate API plan ladder. Treat the three routes as separate budgets unless Speechify's official account or contract terms expressly connect them.

That boundary matters more than the lowest price on any one page. Reader pricing buys a human listening workflow, Studio pricing buys a creator workspace with Studio credits, and API pricing buys programmatic text-to-speech or voice-agent capacity. A buyer who mixes those units can pay for the wrong product or mistake a creator allowance for production API capacity.

Separate the three Speechify routes

The actor is the cleanest routing test. If a person presses play in the Reader, start with Reader pricing. If a creator operates Speechify Studio and exports finished media, start with Studio pricing. If software sends requests, streams speech, or opens agent conversations, start with the SpeechifyAI API plan and its technical limits.

The official pages support a cautious boundary rather than a claim that balances are shared. Reader and Studio link to the API as another Speechify product. Studio's own FAQ says Reader and Studio are separate subscriptions. The API page then defines its own Free, Starter, Pro, Scale, and Enterprise routes, included allowances, prepaid balance, and overage rules.

API plan ladder

The monthly fee covers the listed allowance. Free is a no-card, hard-capped route. Starter, Pro, and Scale are self-serve month-to-month plans; usage above the included allowance draws down a prepaid balance at that plan's published rate, and optional auto top-up can replenish the balance. Enterprise changes the buying process to custom volume, compliance, procurement, and support terms.

Do not choose only by the monthly fee. A higher tier can be cheaper for a large TTS workload because its per-million-character overage falls, while the correct voice-agent tier can depend on included minutes, concurrent calls, phone numbers, support, and operational controls. Compare the total forecast on adjacent tiers, not only the first paid entry point.

Usage-unit table

One subtle distinction prevents bad estimates: billing characters and request-length characters are not the same rule. The API pricing page says whitespace and SSML tags are excluded from the billing count. The API limits documentation says SSML tags do count toward endpoint request limits, with a 2,000-character limit for the speech endpoint and 20,000 characters for the streaming endpoint. Use the pricing definition for cost and the endpoint definition for chunking.

Phone numbers, rate limits, and concurrency are also capacity constraints rather than interchangeable usage units. A plan can have enough included characters yet still be wrong for a bursty workload. Speechify's limits documentation says limits apply per account rather than per API key, so adding keys is not a substitute for choosing adequate account capacity.

Forecast the bill and control spend

Start with a normal month and a spike month. For TTS, estimate all text sent by the product, including alternate voices, regenerated takes, previews, quality checks, retries that create another billable request, and non-production testing. Apply the included allowance first, then model the remainder at the chosen plan's per-million-character rate.

For voice agents, estimate conversation minutes rather than script length. Include test calls, short abandoned sessions, transfers, support peaks, and outbound campaigns where relevant. The published per-minute price includes Speechify's LLM, speech-to-text, text-to-speech, and orchestration components, but buyers should still verify carrier, phone-number, and bring-your-own-carrier terms for their call path.

Prepaid overage and auto top-up reduce the risk that paid production stops at an allowance boundary, but they also create an unattended-spend risk. Speechify documents optional monthly US-dollar limits per API key and a workspace-wide monthly budget. A key that reaches its limit refuses new requests, while the workspace budget can stop new requests and call dispatches across keys and console usage.

Use both controls when separate environments or customers have different risk. Give experiments, staging, and production distinct keys; cap each key; set the workspace budget; and monitor the documented warning events. Speechify notes that billed usage can trail live traffic by a few minutes, so a fast burst may slightly overshoot a cap before enforcement.

Choose the route that matches the actor

Stay with Reader when the job is personal listening and no software integration is required. Choose Studio when a human creator needs voiceover, dubbing, voice changing, cloning, avatar generation, exports, or commercial production rights. Choose the API when a product or backend needs keys, requests, streaming, programmatic generation, or live voice-agent conversations.

Within the API, Free is a controlled starting point for prototypes that can tolerate hard stops. Starter is the first paid route for solo developers and early projects. Pro or Scale becomes more plausible when production volume, concurrent calls, phone numbers, support, or lower overage rates justify the larger base fee. Enterprise is the decision boundary for custom volume, compliance, procurement, SSO, a custom data-processing agreement, or negotiated service terms.

Before paying, verify:

1. The exact Speechify product and workspace that will own the workload.
2. Whether any cross-product entitlement is stated in official account or contract terms; do not infer one from the shared brand.
3. The monthly fee, included characters and minutes, overage rates, prepaid balance, and auto-top-up setting.
4. Billing-character rules separately from endpoint size, rate, and concurrency limits.
5. Phone-number allocation, carrier or BYOC costs, support level, spend caps, and workspace budget.
6. The enterprise commitment and volume terms if self-serve limits do not fit.

The safe purchase rule is simple: buy Reader for listening, Studio for creator output, and SpeechifyAI API capacity for software. Only combine those budgets when Speechify's official terms for the buyer's account explicitly say they connect.