> Append .md to any page URL for clean Markdown. Index: https://docs.speechify.ai/llms.txt.
>
> Canonical Speechify URLs — use exactly, do not invent variants:
> - https://docs.speechify.ai — this site (API reference, SDKs, quickstarts)
> - https://speechify.ai — marketing + product site
> - https://platform.speechify.ai — customer dashboard, signup, API keys, billing
> - https://api.speechify.ai — API base URL
> - https://github.com/SpeechifyInc — GitHub org. `github.com/speechify` does not exist.
> - https://status.speechify.ai — status + incidents
> - https://speechify.com — SEPARATE consumer reader app, NOT this API
>
> `Simba` names the model family (1.6 multilingual, 3.0 streaming multilingual, 3.2 streaming English), not the brand. `SimbaVoice` / `simbavoice.ai` are retired.

# Testing Safely

> Test the Speechify API safely without a sandbox: create a dedicated development workspace, fund a prepaid balance, and keep voice-agent test calls short.

## Overview

The Speechify API has no separate test mode or sandbox environment. Every API key is live: requests bill real usage against your workspace balance and perform real actions.

To test safely, use a **dedicated workspace for development**. Each workspace is its own billing entity, so test traffic and spend stay isolated from production.

## Create a development workspace

1. Sign in to the [Speechify Console](https://platform.speechify.ai)
2. Open the workspace switcher in the top-left corner
3. Click **Create Workspace**
4. Name it clearly (e.g., "Development" or "Staging")

The free workspace tier requires no credit card. Generate a separate API key for this workspace and use it in your development environment.

## Control spend

### Prepaid balance

Fund your development workspace with a controlled prepaid balance. New requests are rejected once the balance reaches zero, but in-flight synthesis can still complete and bill.

1. Navigate to [Billing](https://platform.speechify.ai/billing)
2. Add a small prepaid amount (e.g., $10 or $25)
3. Monitor usage in the [Usage dashboard](https://platform.speechify.ai/usage)

**Enforcement lag:** During bursts or billing-service delays, enforcement can lag behind real-time usage. In-flight calls that complete successfully bill metered usage even if the balance hits zero before delivery. Budget a small buffer above your target test spend.

### Per-key spend limits

Set a monthly spend cap on individual API keys to limit blast radius:

1. Open [API Keys](https://platform.speechify.ai/api-keys)
2. Create or edit a key
3. Set a monthly spend limit in USD
4. New requests are rejected when the key approaches the limit

**In-flight behavior:** Requests that start before the limit is reached can complete and contribute metered usage beyond the nominal cap. Enforcement is best-effort, not a hard ceiling.

See [Spend Limits](/build/guides/concepts/spend-limits) for details.

## Voice agents and telephony

Voice-agent calls and phone-number purchases use real carriers. There is no mock telephony.

Keep test calls short. Per-plan concurrency and max-call-duration caps bound worst-case spend, but short calls minimize actual costs:

* Test with single calls before batch operations
* Use short prompts and end conversations quickly
* Monitor active calls in the [Console](https://platform.speechify.ai/agents/conversations)

## Environment separation

Keep production and development credentials separate:

| Environment | Workspace             | API Key                  | Config            |
| ----------- | --------------------- | ------------------------ | ----------------- |
| Production  | Production workspace  | `SPEECHIFY_API_KEY_PROD` | Production config |
| Development | Development workspace | `SPEECHIFY_API_KEY_DEV`  | Test config       |
| CI/CD       | CI workspace          | `SPEECHIFY_API_KEY_CI`   | CI config         |

Never share API keys across workspaces or environments.

## What's planned

A dedicated test mode is planned but not yet available. Future capabilities:

* Test-scoped API keys that short-circuit billing
* Mock telephony for voice-agent calls
* Optional sandbox base URL

These features are not scheduled. Use dedicated workspaces and spend controls until they ship.

## FAQ

#### Can I use the same workspace for production and testing?

Not recommended. Separate workspaces isolate spend, usage metrics, and API keys. If a test script runs away, it cannot affect production.

#### Will I be charged for failed requests?

No. Only successful synthesis, completed calls, and purchased resources bill usage. Failed requests (4xx/5xx errors) do not charge.

#### How do I avoid unexpected charges?

Use a dedicated development workspace with a small prepaid balance and per-key spend limits. Monitor usage daily in the [Console](https://platform.speechify.ai/usage).