High-volume authenticity is a workflow problem, not a feature flag. Agency editors sweeping a week of freelancer drafts, content platforms running moderation rewrites, and SEO teams rewriting a thousand-post archive all hit the same wall the moment they paste piece number eleven into a single-shot UI. The browser tab is the bottleneck, not the model. TextSight Business solves bulk on two surfaces in one tier: a Bulk Rewrite dashboard that takes paste, drag-and-drop, and ZIP upload across eleven file formats, and a REST batch endpoint at api.textsight.ai/humanizer/v1/batch that accepts an array of documents and returns an array of rewrites. Workspace seats, audit log, and a flat 150,000-word monthly allowance round out the agency shape.
A solo blogger rewriting one post a week does not need bulk. Bulk becomes the right answer when the unit of work is a stack of drafts that has to be done before Tuesday and the per-piece overhead has to be near zero for the workflow to scale.
A content agency with five writers and six active clients processes 25 to 40 pieces a week, each one needing a defensible authenticity step before invoicing the client. The agency wants one billing relationship, five seats, and a consolidated usage view. The Bulk Rewrite UI drains the week's drafts in one queue; the workspace audit log tags each run with the client label so deliverables stay separable. Gross margin on the loop depends on it being cheap, which is why credit-based pricing breaks the model.
A user-generated content platform screens every submission with the detector, then offers a rewrite-and-resubmit suggestion on the flagged ones rather than a hard reject. That is the batch endpoint pattern: the moderation queue posts an array of flagged drafts, gets back an array of rewrite suggestions, and surfaces them in the contributor UI. The flat monthly allowance protects the platform from credit-pack runaway when traffic spikes.
An SEO team auditing a thousand-post archive for AI-written content that pre-dates the team's quality bar needs every flagged piece rewritten in one push. The script reads the sitemap, runs the detector, batches the flagged pieces through the rewrite batch endpoint overnight, and writes the results back. Flat allowance fits the steady nightly volume; credit pricing would burn the budget on this shape. The audit log records the entire run for compliance and rollback.
Open the dashboard, click Bulk Rewrite, and feed the queue with paste, drag-and-drop, or ZIP upload. The unified extractor handles 11 formats via officeparser. Each item lands with the original word count, the rewrite, and the Authenticity Score side by side.
Paste plain text for fast one-offs inside a bulk run. Drag-and-drop individual files for .docx, .pdf, .txt, .md, .rtf, .odt, .epub, .html, .pptx, .xlsx, and .csv. Upload a ZIP archive when an agency wants to push a whole client folder in one shot. Mixed-format ZIPs are fine. The unified /api/file/extract route is the same officeparser pipeline that powers the Bulk Scan UI for the detector, so format coverage is consistent across detect and rewrite.
Pick Standard for routine drafts and Maximum for heavily templated or LLM-stamped prose where Standard leaves residual signals. Mode is a per-batch setting; the queue applies it uniformly across the run. A second pass on Maximum is one click for items that come back under Authenticity Score 75.
Each item displays the original word count, the rewrite, the chosen mode, and the Authenticity Score. Sort by score to find the rewrites that need a second pass. Click any row for the full reading view with before-and-after side by side. Export per-piece or per-batch; bulk CSV export is available via the API today and is rolling into the dashboard.
The queue keeps processing while the editor tab is closed because the job runs server-side once submitted. Return later and the results are in your history. The workspace history is shared on Business so a colleague can pick up the run if you handed it off.
When the trigger is another system rather than a human reviewer, the batch endpoint is the right call. POST an array of documents in one request, receive an array of rewrites in the same order. Cheaper and faster than fanning out one synchronous call per document.
POST JSON to api.textsight.ai/humanizer/v1/batch with an Authorization: Bearer YOUR_API_KEY header. The body carries a documents array, each entry with a text field, plus a top-level mode value of standard or maximum. Optional return flags request the Authenticity Score and before-and-after detector scores. Authentication is the same key issued from the /api-keys page in the dashboard, scoped to the issuing account and the team workspace it belongs to.
The response is JSON with a results array in the same order as the request. Each entry carries the rewritten rewrite, the mode echoed back, the humanizationScore as a 0 to 100 measure, and optionally the before and after detector scores when the return flag was set. Failures on individual items return a per-item error object instead of rewritten, so a partial batch still ships the successful pieces.
The endpoint accepts up to 50 documents per call by default. Above that, split into multiple calls; the batch is the unit of TLS and warm-up amortisation, and 50 is the sweet spot for both quality and throughput. End-to-end latency on a 50-document batch of 1,000-word pieces runs in the 30 to 90-second band depending on mode and warm state. Items return as they finish; the response holds open until the slowest item lands.
HTTP 429 with a Retry-After header signals you have hit the monthly allowance. 401 means the key is missing or revoked. 5xx is rare and transient: back off exponentially, retry up to three times, then route to the audit log for manual triage. Every batch is recorded in the workspace audit log with timestamp, requester, item count, and total word count consumed.
Where the queue, the batch endpoint, and the workspace each earn their keep. Use the API when the trigger is another system; use the UI when a human reviewer is in the loop.
An agency editor receives ten to forty drafts a week from freelancers across six clients. The editor opens Bulk Rewrite, drags the week's drafts in, runs Maximum mode, and exports the rewrites to the project tracker tagged by client. Workspace seats let two editors share one billing and one history. The audit log becomes the defensible record that every invoiced piece went through authenticity, which is the artifact procurement and client review both ask for.
A user-generated content platform screens every submission with the scan endpoint. Flagged pieces above the platform threshold are not rejected outright; they are piped through the rewrite batch endpoint as a rewrite suggestion surfaced in the contributor UI. The contributor accepts, edits, or discards. The batch endpoint handles asynchronous spikes when the queue fills. Flat monthly allowance keeps the cost predictable even when traffic doubles unexpectedly.
An SEO team or editorial group audits a thousand-post archive for AI-written content that pre-dates the team's quality bar. A nightly job reads the sitemap, runs each post through the detector, batches the flagged pieces through api.textsight.ai/humanizer/v1/batch on Maximum, and writes the rewrites into a staging table for review. The audit log records every change. Pricing math at this shape strongly favours flat allowance over per-word credits.
Bulk authenticity at agency scale needs more than a queue. It needs a workspace where five writers, two editors, and one owner all see the same history; an audit log that names who rewritten what and when; and client-level tagging so a single account can run six client books without mixing them.
Business ships five workspace seats. The owner invites writers and editors by email; each accepts and lands under their own login with the team's shared allowance. Roles cover owner, admin, and member with role-based access on billing and seat management. Pending invites can be revoked, active seats can be removed, and the invite flow handles the edge cases (brand-new owner first invite, last pending invite acceptance) that the May 2026 team-management audit caught and fixed.
Every rewrite batch, every scan, every seat change is recorded in the workspace audit log with timestamp, actor, and payload size. The log is the artifact that compliance, finance, and client review all ask for, and it is the difference between five personal accounts and a real team workspace. Filter by date range, by actor, or by run type for monthly invoicing and quarterly compliance reviews.
Each bulk run accepts an optional client label that propagates into the audit log and the export filename. An agency running six client books inside one account keeps the deliverables separable without spinning up six logins. The same pattern works for an editorial team running multiple sections or a content platform tagging by tenant.
Bulk rewrite history is shared workspace-wide, so a colleague who picks up a run mid-flight sees the same queue and the same per-piece outputs. Personal history is preserved as well; the shared view aggregates everyone's work without erasing the per-seat trail. This is the loop that lets a team hand off work without losing the audit chain.
At bulk scale, pricing model matters more than headline price. Per-word and per-credit pricing punish heavy users. Flat monthly allowance flattens the curve and keeps finance happy.
Per-piece cost math at three volume bands. At 50 rewritten pieces a month, credit-based pricing and flat allowance look comparable; at 500, flat allowance pulls ahead; at 5,000, flat allowance is roughly an order of magnitude cheaper because the monthly fee is fixed while credit packs scale linearly with usage.
Free, Starter, and Pro cover individual writers and editor UIs. Business adds the Bulk Rewrite UI at scale, the REST batch endpoint, five workspace seats, and the audit log. Full details on the pricing page.
Billed $89.88/year, Save $30
Billed $179.88/year, Save $60
Billed $359.88/year, Save $120
Yearly billing saves 25%. View full pricing →
REST and SSE streaming docs for developers wiring authenticity into a product backend or pipeline.
See API docs →The detection side of the same workflow. Bulk Scan UI, batch endpoint, and workspace audit log on Business.
See bulk detection →Agency workflow with Bulk Rewrite, multi-client tagging, and the workspace audit log for invoicing.
See agency workflow →Free, Starter, Pro, Business. Bulk UI and batch API live on Business at $29.99/mo yearly.
See pricing →Bulk Rewrite UI with 11-format upload and ZIP. REST batch endpoint at api.textsight.ai/humanizer/v1/batch. Five workspace seats with shared history and workspace-wide audit log. Flat pricing that gets cheaper per-piece as volume rises.