> ## Documentation Index > Fetch the complete documentation index at: https://docs.stealthgpt.ai/llms.txt > Use this file to discover all available pages before exploring further. > ## Agent Instructions > Start with https://docs.stealthgpt.ai/api-reference/quickstart.md for REST setup and https://docs.stealthgpt.ai/api-reference/models.md for Super, Standard and Lite model selection. > For existing drafts, use the Writer & Humanizer endpoints; for research-to-draft generation, use the Stealth Agent guides. Choose polling, webhook callbacks or MCP with https://docs.stealthgpt.ai/api-reference/workflow-patterns.md. > For bulk humanization and editorial review, use https://docs.stealthgpt.ai/api-reference/recipes/humanize-review.md. For fixed links, citations or other approved spans, use https://docs.stealthgpt.ai/api-reference/text-preservation.md and replace each protected span with a marker in your application. > For assistant setup and verified client status, use https://docs.stealthgpt.ai/api-reference/integrations/mcp.md. Consult the Pricing and Data Handling references for API entitlements, word accounting and processing details; web subscriptions and API plans are separate. # StealthGPT API data handling > Plan an API integration with clear information about request history, processing providers, async callbacks, account controls and policy references. Build your StealthGPT API workflow around the data you submit, the results you collect, and the review process you control. This guide explains the current data flow for humanization, writing, detection and Stealth Agent so you can prepare an integration or procurement review. ## Requests, results and account history REST requests use the **`api-token`** header to identify your StealthGPT account. Async status lookup checks that the run belongs to that account. The [API dashboard](https://www.stealthgpt.ai/api-dashboard) provides account-scoped history with request text, returned content and usage information. | Workflow | Information handled | | - | - | | Writer and humanizer | API history records include submitted text or writing prompts, returned text, model and usage/billing metadata. | | Async humanization | A run record tracks status, progress, errors and the completed result's history record. The source text and selected options are passed to background processing. | | AI detection | The submitted text, detection result and usage/billing metadata are recorded. | | Stealth Agent | The run stores its brief, status and progress information; generated documents are associated with the account. Completed, billed API runs appear in API history. | | MCP | The connected StealthGPT account owns the run. Humanization uses the async processing path; subscription and API-credit billing can write to different account-history records. | Keep your own copy of completed results and run IDs. Your publishing system should retain the reviewed version of a document and its approval record. The [humanization review recipe](https://docs.stealthgpt.ai/api-reference/recipes/humanize-review.md) demonstrates this handoff without publishing directly to a CMS. See [async runs](https://docs.stealthgpt.ai/api-reference/endpoints/stealthify-runs.md), [Stealth Agent](https://docs.stealthgpt.ai/api-reference/stealth-agent/overview.md) and [MCP billing](https://docs.stealthgpt.ai/api-reference/integrations/mcp.md#billing) for each workflow's response fields and account behavior. ## Processing follows your chosen workflow StealthGPT processes request content using hosted inference and supporting services. Current implementation paths include: | Service | Processing role | | - | - | | Modal | Hosted humanization inference for Super and Standard. | | OpenRouter and its routed model providers | Lite humanization, Stealth Agent drafting and optional image generation, plus model-backed translation and selected research, quality or formatting utilities. | | OpenAI | Embedding-based similarity checks and preservation workflows that use embeddings. | | Trigger.dev | Background execution of async jobs, including submitted text and selected options. | | GPTZero | The API detection integration submits document text to its detector. | | Exa | Stealth Agent web research sends generated search queries and retrieves source material. | The route depends on the model, requested features and fallback behavior. Stealth Agent can add research, fact checking and image generation to the workflow. Use the [model guide](https://docs.stealthgpt.ai/api-reference/models.md) and [Stealth Agent controls](https://docs.stealthgpt.ai/api-reference/stealth-agent/guide.md) to choose the capabilities you need. For a processor inventory or a specific processing-region requirement, include your intended workflow in a [support request](https://www.stealthgpt.ai/support). ## Callbacks send results to your endpoint Providing `webhookUrl` opts an async run into a terminal callback. That endpoint receives the completed result and billing fields, or the terminal error payload. Choose an endpoint under your control and apply your own access, logging and retention rules there. Set `webhookSecret` to receive an HMAC-SHA256 signature. Verify the timestamp and raw request body, deduplicate by run ID, and retain polling as a reconciliation path. The [webhook reference](https://docs.stealthgpt.ai/api-reference/endpoints/stealthify-runs.md#webhook-callbacks) documents signing and delivery behavior; the [review recipe](https://docs.stealthgpt.ai/api-reference/recipes/humanize-review.md#optional-signed-webhooks-with-polling-reconciliation) includes a receiver example. ## Retention and deletion requests The [Privacy Policy](https://www.stealthgpt.ai/privacy) sets out personal-information retention terms: retention as needed for its stated purposes, a three-month post-termination limit for those purposes, and provisions for legal requirements and backup archives. Follow its contact instructions to request access, updates or deletion of personal information. An API run ID is a retrieval reference, rather than a per-request retention setting. Keep completed results in your own system for the lifecycle your application requires. For an agreed retention requirement or account-specific deletion request, use the policy's contact route or [support](https://www.stealthgpt.ai/support). ## Terms, commercial deployments and billing Review the [Terms of Service](https://www.stealthgpt.ai/tos), particularly the intellectual-property, permitted-activity, API-policy and contribution-license sections, for your planned use. Use the documented REST or MCP interfaces for integrations, and describe any commercial-permission or contract requirement in a support request. For implementation planning, [API pricing](https://docs.stealthgpt.ai/api-reference/pricing.md) explains plan usage and pay-as-you-go. Humanization counts input plus output billable words: Standard and Lite use a 1× multiplier and Super uses 2.5×. MCP's account billing selection is documented [separately](https://docs.stealthgpt.ai/api-reference/integrations/mcp.md#billing). The API dashboard and [usage settings](https://www.stealthgpt.ai/account/usage) provide the corresponding account controls. Data-flow summary reviewed November 7, 2026. The linked policies provide the applicable policy terms. This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.