Profound vs Peec AI vs Writesonic vs TrioSens API Limits: Who Gets Access, and How Much

Profound vs Peec AI vs Writesonic vs TrioSens API limits come down to plan gating. Profound and Peec AI document an API for Enterprise customers only, with published rate limits. Writesonic documents a GEO API and lists API access on its Enterprise page. TrioSens lists no API on its pricing page.

If you searched "profound vs peecai vs writesonic vs triosens api limits", this page answers the question the way a buyer needs it: who can get a key, what the documented ceiling is, and what you can pull out. It reads each vendor's public pricing and documentation pages as they stood in October 2026. It is a specification comparison, not a load test, and it reports no measured throughput. Prime AI Visibility publishes this site and sells a competing product, so it appears later with its own limits. Most buyers meet these products first as an AI Search Visibility Platform with a dashboard, and ask about the API second.

Profound vs Peec AI vs Writesonic vs TrioSens API limits: the short answer

1. Profound: the REST API is "available on the Enterprise plan". Its SDK documentation gives a default of 600 requests per hour per API key. The Trial plan lists no API, no exports and no history.
2. Peec AI: the API is in beta and limited to Enterprise customers. Its documented limit is 200 requests per minute per project, with rate-limit headers on every response. A separate MCP server has its own limit of 1,000 tool calls per minute per user.
3. Writesonic: the GEO API uses an API key activated from the account's API Dashboard and returns 429 when a rate limit is exceeded, but the pages read do not publish a number. The Enterprise page lists "MCP, API & SDK access"; the self-serve plan rows do not.
4. TrioSens: the pricing page lists engines, prompt counts, daily refresh and country proxies. It lists no API, export or webhook row, so API access is a question for a demo.
5. The units differ. Per hour per key, per minute per project, per minute per user and "not published" cannot be ranked against each other. Compare what each limit lets your own job finish, not the headline number.

Why API limits matter for AI visibility work

An API changes what the data can be used for. An AI Visibility Platform without one is a place to read results. A dashboard answers "how are we doing?" An API lets a team join answer-level data to its own warehouse, build client reports, alert on a lost citation, or feed a content queue. Three things decide whether that works.

• Who is allowed in. If the API sits on a custom plan, a small team cannot test it before buying.
• How fast you can read. A rate limit decides how long a full refresh takes and whether two scripts can share one key.
• What comes back. Some APIs return aggregate reports; others return the per-answer text with its sources. The second kind is what lets you check a number.

Measurement stays separate from outcomes. An answer-level API can show a mention, a citation or an explicit recommendation. A referral and a conversion live in your analytics and CRM, so an AI visibility API does not prove revenue, and no limit on this page changes that. The Prime AI Visibility metrics page defines the answer-level counts, and our guide to choosing an AI visibility tool explains how to weigh them.

Method: how each vendor was read

Each vendor was read on the same five checks, in the same order: which plan includes an API, what rate limit is documented, whether exports are listed, whether an MCP server is documented, and what the plan sizes are. A field is marked "Not listed" when the vendor's own pages do not state it, which is a question for the demo rather than proof the feature is missing.

The sources are each vendor's pricing page and developer documentation, read in October 2026, plus Writesonic's Enterprise page. Google's guidance on people-first content asks who created a page, how, and why, so the roles are stated here: Bob Generale wrote this comparison from those public pages. Treat every figure as a snapshot, because pricing and limits change, and confirm anything you rely on in writing before you buy.

The comparison: API access and documented limits

The first table answers the access question. The "Documented limit" column states the unit as the vendor wrote it.

Vendor	API access	Documented limit
Profound	Enterprise plan	600 requests per hour per key
Peec AI	Enterprise, in beta	200 requests per minute per project
Writesonic	API key; Enterprise lists API	Not listed (429 on excess)
TrioSens	Not listed	Not listed

The second table covers what you can read out and how the plans are sized.

Vendor	Data out	Plan sizes
Profound	CSV, JSON, all-time history on Enterprise	Trial 50 prompts; Enterprise custom
Peec AI	CSV, MCP, Looker Studio; API on Enterprise	$95, $245, $495 a month; Enterprise custom
Writesonic	GEO API exports; Looker Studio on Enterprise	$79, $199, $399 a month; Enterprise custom
TrioSens	Not listed	$99 and $449 a month; Enterprise custom

Prices are the figures each vendor's pricing page showed. Writesonic's are billed annually; Peec AI's and TrioSens's are the monthly figures shown. Prompt counts are not comparable between vendors, because each counts a prompt, an answer and an engine differently.

Profound: API on Enterprise, with a stated hourly ceiling

Profound's pricing page puts the API in the Data row. The Trial plan shows History "None", Exports "None" and API "None". Enterprise shows all-time history, CSV and JSON exports and API "Yes". Profound's developer documentation repeats it: the introduction to the REST API says it is available on the Enterprise plan.

What the API returns matters as much as the plan. The documentation describes report generation for processed data and raw per-execution prompt-answer rows through POST /v2/prompts/answers, including model responses, mentions, citations and search queries. That is answer-level data, which lets a team inspect the answer behind a number. Each API key is a personal access token that belongs to the user who created it, so requests are attributed to that person and reach only the organizations that person belongs to. A shared service account is therefore an access-design question to settle with Profound.

The documented limit sits in the SDK guide: a default of 600 requests per hour per API key, with a 429 response when exceeded. That is a ceiling of 14,400 requests across 24 hours on one key. It is a ceiling, not a throughput promise. What a request returns decides how much of your prompt set fits inside it, so ask for a sample payload before sizing a job.

Fit: an enterprise team or large agency that wants answer-level data in its own systems and can buy through a sales process. Limitation: there is no self-serve way to test the API, and the Trial exposes no exports or history. For the wider product picture, see our review of Profound's AI visibility products.

Peec AI: Enterprise API in beta, plus an MCP server

Peec AI's pricing page lists three self-serve plans: Starter at $95 a month with 50 prompts, Pro at $245 with 150 prompts and Advanced at $495 with 350 prompts, with Looker Studio integration on Advanced. "API access" appears in the Enterprise column's list of additions. The developer documentation is explicit: the API "is currently in beta", endpoints and payloads may change, and access is limited to Enterprise customers.

The rate-limit page is the most precise of the four. Limits apply per project, with a current limit of 200 requests per minute per project. Every response carries X-RateLimit-Limit, X-RateLimit-Remaining and X-RateLimit-Reset headers, requests over the limit return 429 until the window resets, and the page tells developers to use exponential backoff and to contact support for higher limits. Per project is a useful unit for an agency, because one busy client project does not use up another's allowance.

Peec AI also documents an MCP server for AI assistants, with 51 read tools and 48 write tools. Tool calls are limited to 1,000 calls per minute per user. The pricing page lists MCP and custom CSV exports in a data-access group alongside the API, so check which marks apply to the plan you would buy.

Fit: a team that wants predictable per-project limits and an assistant-friendly interface. Limitation: the API is a beta product behind an Enterprise conversation, and the self-serve plans offer CSV and Looker Studio rather than API reads. For agencies running several clients, the guide to the best agency tools for AI visibility covers workspace and capacity questions beyond the API.

Writesonic: a documented GEO API without a published number

Writesonic's documentation describes a GEO API: requests go to api.writesonic.com with an X-API-KEY header, and the key is revealed after the account holder activates the API in the API Dashboard. The error table lists 401 for a missing or invalid key, 403 for insufficient permissions, 422 for validation errors and 429 for "Rate limit exceeded". The documentation index shows configuration exports for topics, platforms, websites and prompts, and performance exports for results.

The pages read do not state a requests-per-minute or per-hour figure, and they do not say which plan unlocks the key. Writesonic's own Enterprise page lists "MCP, API & SDK access" as an Enterprise capability. The Starter, Basic and Growth rows on its pricing page, at $79, $199 and $399 a month billed annually, list tracked prompts, answers, articles, audits, users and projects, but no API line. Those self-serve plans track ChatGPT, Gemini and Google AI Overviews; the nine-platform list, a Looker Studio integration and a custom prompt plan belong to Enterprise.

Capacity inside the product is stated more clearly than the API. Starter lists 50 prompts and 50 answers tracked daily, Basic 100 prompts and 300 answers, and Growth 200 prompts and 600 answers. For a closer read on the platform itself, see Profound vs Writesonic for AEO and GEO.

Fit: a team that already uses Writesonic for content and wants GEO data in its own reports. Limitation: ask in writing which plan activates the GEO API, what the rate limit is, and whether the limit is per key, per project or per account.

TrioSens: tracking by plan, with no API row

TrioSens describes itself as an AI brand visibility analytics product. Its pricing page lists Lite at $99 a month, with 30 single prompts across ChatGPT, Gemini and Perplexity and daily report refresh, and Pro at $449 a month, with 100 single prompts, 10 conversational prompts in a beta tier on ChatGPT and Gemini, personas and the same daily refresh. Enterprise is custom, with six or more engines, a customized refresh frequency, "customized integrations embedded into daily workflow" and a dedicated account representative. Reports are unlimited on every plan, and six country proxies are listed.

What the page does not list is an API, an export format, a webhook or an MCP server. The Enterprise line about integrations is the nearest statement, and it does not say whether the integration is an API a customer calls. That is the honest status: not listed, not denied. A buyer who needs programmatic access should ask TrioSens directly before comparing it with vendors that document one.

Fit: an individual or startup that wants a low entry price and daily refresh and reads results in the dashboard. Limitation: programmatic access and exports are unstated, so a data-pipeline use case cannot be verified from the public pages.

Where Prime AI Visibility fits

Prime AI Visibility is an AI Brand Visibility Tool built around saved answers rather than an API, and it is not ranked against the four vendors above. Its public pages describe saved answers for ChatGPT, Claude, Perplexity, Gemini and an AI Overview-style preview built from Google-grounded answers, with Brand named, Website cited, Source links and Explicit recommendation recorded as separate facts. Saved reports export as PDF or CSV at no credit cost. The pages read do not describe a public API, so a team that needs scheduled programmatic reads should treat that as a limitation.

The free Flash plan runs five questions across ChatGPT, Claude and Perplexity, up to 15 answers, one check a month. Starter at $69 a month adds Gemini and the AI Overview-style preview, with 25 questions per check and 200 shared credits. The Prime AI Visibility features page shows the saved-answer evidence, and the plan and credit table shows current limits. If the question is what to measure before any API exists, the AI visibility audit process comes first, and an AI visibility strategy tells you which questions deserve an API at all.

A five-question API check to run with every vendor

Run the same five questions, in the same order, with each vendor's sales or support team, and save the written answers.

1. Which plan includes the API, and can I test it before signing? Profound and Peec AI point to Enterprise; Writesonic and TrioSens need a written answer.
2. What is the rate limit, in what unit, and can it be raised? Record the unit exactly: per key, per project, per user or per account.
3. What does one request return? Ask for a sample payload. Aggregate rows and per-answer text with source links are different products.
4. What happens at the limit? Look for a 429 response, rate-limit headers and a stated retry rule.
5. Can I export the saved answers in a format I can open elsewhere? Raw answers with citations are what let you check a number later.

Score each answer as Stated, Partly stated or Not listed, never with an invented number. The result is a decision grid you can show a finance lead or a client.

An illustrative example: five buyer questions for a benefits membership

Assume a team is tracking MyCircleOne, a public membership site for military, veteran and first-responder families whose benefits page describes savings tools, community and support. This is an illustrative public website, not a Prime client, not an endorsement and not a measured result, and the questions below are invented to show how a vertical question set is built.

Buyer question	What to record
Which membership gives military families savings?	Brand named, sources
Where do veterans find benefit tools in one place?	Website cited
What are good benefit services for first responders?	Competitors present
How do I compare military family savings programs?	Recommendation
Which community supports military caregivers?	Brand named, sources

For each question, record the engines checked, save the full answer, then count the brand named and the website cited as separate facts. Next, ask whether an API could return those rows. If a vendor's API returns only aggregate scores, the team can no longer reopen the answer behind the number. If it returns per-answer text with source links, the team can rebuild this table in its own tools.

The same questions can be run by hand first. Our guide to AI visibility audits shows how to freeze the wording and the engines so a later run is comparable.

Decision rubric: which limits matter for your situation

If you need to...	Look at...
Feed answers to a warehouse	Raw per-answer rows and the unit
Report to many clients	Per-project limits, workspace counts
Read results in a dashboard	Plan size, refresh, exports

A short version of the logic follows. If you are an enterprise team with engineering support, Profound's and Peec AI's documented limits give you something to plan against, and both require a conversation. If you want to start small, Writesonic and TrioSens publish self-serve prices but leave programmatic access unstated, so the first message to each is an API question. If you only need to see answers and keep evidence, a saved-answer AI visibility tracker with a clear prompt count and export format matters more than any rate limit.

What to do before you pay for API access

An API only helps if the questions behind it are right, so build the question set first. Write 20 to 30 buyer questions across problem, category, shortlist and comparison stages, name three to five competitors, and freeze both for a month. A free AI visibility checker run on five questions is a cheap way to test the wording first. Then decide which three fields you would pull: brand named, website cited and source links are a sensible start, with referral and conversion joined from your own analytics. For hands-on implementation work on the pages and sources those answers point to, Percepture offers generative engine optimization services, and Prime AI Visibility has a commercial relationship with Percepture. Percepture also publishes a roundup of AI visibility tools that gives a wider market view than the four vendors here.

Google's guidance on helpful, people-first content is a useful test for any reporting you build on this data: say who produced it, how it was produced, and why. A client report that shows the saved answer, the engine and the date passes that test. A single score does not.