Otterly AI API, MCP, exports, webhooks and integrations
A source-checked guide to Otterly AI REST API and MCP access, quotas, exports, Looker Studio, Agent Analytics ingestion, webhook limits, and the Trakkr comparison.
Quick answer
Does Otterly AI have the API, MCP, exports, webhooks, and BI connections this team needs?
Otterly AI has a documented REST API, an OAuth MCP server, CSV, JSON and PDF exports, and a Looker Studio connector. API and MCP start on Standard with 2,000 monthly requests and rise to 5,000 on Premium. Agent Analytics accepts inbound log events by webhook. A native outbound event-webhook product is not publicly documented, so alerts and workflow delivery require the API or another automation layer.
Key facts and evidence
- REST API plans
- Standard, Premium, and CustomEvidence: Otterly AI public API guide, Otterly AI pricing and plan comparison
- API authentication
- Admin-managed Bearer keys with read-only or read-write accessEvidence: Otterly AI public API guide
- MCP transport
- Hosted streamable HTTP with OAuth 2.0Evidence: Otterly AI MCP server guide
- Exports
- CSV, JSON, and PDF paths documented on all plansEvidence: Otterly AI data and report exports
- Looker Studio
- Available on Standard, Premium, and CustomEvidence: Otterly AI Looker Studio connector, Otterly AI pricing and plan comparison
- Webhooks
- Inbound crawler-log ingestion documented; outbound events not publicly documentedEvidence: Otterly AI Agent Analytics setup, Otterly AI pricing and plan comparison, Otterly AI public API guide, Otterly AI MCP server guide, Otterly AI data and report exportsNot publicly verified. Otterly suggests building alerts through its API and automation tools, but that is not the same as a documented outbound event-delivery service. Confirm private or planned webhook support directly.
Otterly AI integration path by buyer job
| Integration job | Otterly AI route | Plan or limit | Buyer check |
|---|---|---|---|
| Programmatic reads | REST API for reports, mentions, coverage, citations, share of voice, sentiment, and prompts | Standard 2,000 monthly requests; Premium 5,000; Custom negotiated | Test pagination, historical windows, workspace scope, and monthly consumptionEvidence: Otterly AI public API guide, Otterly AI pricing and plan comparison |
| Programmatic writes | Read-write keys can run GEO audits and create or delete prompts and tags | Same plan quota plus a 2,000-request rolling five-minute rate limit | Do not assume every dashboard object has a write endpointEvidence: Otterly AI public API guide |
| AI assistants | OAuth MCP exposes reports, prompts, responses, citations, recommendations, and audits | 2,000 monthly MCP requests on Standard; 5,000 on Premium | Confirm client compatibility and write permission before automationEvidence: Otterly AI MCP server guide, Otterly AI public API guide, Otterly AI pricing and plan comparison |
| Business intelligence | Looker Studio can select a workspace report, engine, country, and tag | Standard, Premium, and Custom | Validate fields and refresh behavior against the planned dashboardEvidence: Otterly AI Looker Studio connector, Otterly AI pricing and plan comparison |
| File delivery | CSV prompts and citations, CSV or JSON responses, PDF reports and audits | Published on all plans | Choose exports when API automation is unnecessary or quota-sensitiveEvidence: Otterly AI data and report exports |
| Event webhooks | Inbound Agent Analytics log webhook; no public outbound event product found | Agent Analytics begins on Standard | Ask about events, signatures, retries, logs, and delivery guaranteesEvidence: Otterly AI Agent Analytics setup, Otterly AI pricing and plan comparison, Otterly AI public API guide, Otterly AI MCP server guide, Otterly AI data and report exportsNot publicly verified. Otterly suggests building alerts through its API and automation tools, but that is not the same as a documented outbound event-delivery service. Confirm private or planned webhook support directly. |
An inbound log-ingestion webhook is not an outbound notification webhook. The direction of data flow matters when designing alerts and automations.
What can a team read and write through the Otterly AI API?
Read access covers brand mentions, brand coverage, domain citations, share of voice, sentiment, and prompt-level results. Admins create workspace-scoped keys and choose read-only or read-write permission.
The documented write surface is narrower: GEO audits plus creation and deletion of prompts and tags. Standard includes 2,000 monthly API requests, Premium 5,000, and the technical rate limit is 2,000 requests in a rolling five-minute window.
Can ChatGPT, Claude, Cursor, or another assistant query Otterly AI through MCP?
Yes. Otterly publishes a hosted streamable-HTTP MCP endpoint with OAuth 2.0 rather than API-key authentication. Its compatibility list includes ChatGPT, Claude, Claude Code, Cursor, n8n, and other clients that support the transport.
The tool surface includes visibility KPIs, prompt results, raw answers, citations, recommendations, GEO audits, and permission-gated prompt, tag, and audit writes. Plan request quotas still apply.
Can Otterly AI data leave the product without custom engineering?
Yes. All plans publish several direct export paths: CSV for prompts, citations, and recommendations; CSV or JSON for raw answers; and PDFs for brand reports and GEO audits.
Standard and higher plans add Looker Studio. That connector supports live dashboards and can scope the selected data by report, engine, country, or tag.
Does Otterly AI send outbound webhooks when visibility changes?
No native outbound event product is documented in the public material checked. Otterly does document a webhook in Agent Analytics, but customers post crawler and referral log events into Otterly through that endpoint.
Its API guide suggests using API data with tools such as Zapier, Make, or Slack to build alerts. Buyers needing native push delivery should confirm event types, signing, retries, failure logs, and plan entitlement directly.
When is Otterly AI or Trakkr stronger for integrations?
Otterly is stronger when a Standard buyer needs REST API access, because Trakkr gates its REST API to Scale and Enterprise. Otterly also documents a broad API read surface and permission-gated writes.
Trakkr is stronger when native outbound webhooks, Slack and task destinations, Google Sheets, or paid-plan MCP are central. Both products publish Looker Studio and MCP, but their plan gates and automation direction differ.
Evidence and method
API scope is documented beyond a feature badge
Otterly publishes data categories, key permissions, plan quotas, authentication, rate limits, and a bounded write surface for implementation review.
Evidence: Otterly AI public API guide, Otterly AI pricing and plan comparisonMCP has a current transport and tool map
The MCP guide names OAuth 2.0 streamable HTTP, compatible clients, data categories, writes, common errors, and plan usage controls.
Evidence: Otterly AI MCP server guideExports are documented by format and plan
The export guide identifies CSV, JSON, and PDF routes and separates all-plan file exports from the higher-plan Looker Studio connector.
Evidence: Otterly AI data and report exports, Otterly AI Looker Studio connector, Otterly AI pricing and plan comparisonWebhook direction is explicit
Agent Analytics documents customers posting log batches into Otterly, which prevents an inbound connector from being mistaken for outbound event delivery.
Evidence: Otterly AI Agent Analytics setup, Otterly AI pricing and plan comparison, Otterly AI public API guide, Otterly AI MCP server guide, Otterly AI data and report exportsNot publicly verified. Otterly suggests building alerts through its API and automation tools, but that is not the same as a documented outbound event-delivery service. Confirm private or planned webhook support directly.How we checked this page
We separated file exports, REST API, MCP, Looker Studio, inbound log ingestion, outbound events, and third-party workflow destinations before checking each entitlement and data direction.
- 1. Read Otterly AI's current API, MCP, export, Looker Studio, Agent Analytics, and pricing documentation.
- 2. Mapped authentication, permissions, quotas, formats, data direction, and plan gates to the exact buyer workflow.
- 3. Compared Trakkr only from its current official integration directory and API reference.
- Limitation: We did not authenticate to the Otterly API or MCP server, measure connector latency, or inspect negotiated Custom quotas.
- Limitation: The outbound-webhook finding is limited to public documentation and should be reconfirmed for private Enterprise capabilities.
When is Otterly AI or Trakkr the better integration fit?
Choose Otterly AI when a Standard-level REST API, broad direct exports, OAuth MCP, or its Agent Analytics log-ingestion webhook matches the workflow. Its API entitlement arrives below Otterly Premium and includes documented read and selected write operations.
Choose Trakkr when native outbound webhooks, Slack and task tools, Google Sheets, or MCP on every paid plan matter more. Trakkr's REST API starts on Scale, so the correct choice depends on destination, direction, and plan rather than a simple integration count.
Yes. Standard includes 2,000 monthly API requests, Premium includes 5,000, and Custom plans can negotiate limits. Lite does not publish API access.
Yes. Its hosted server uses OAuth 2.0 and streamable HTTP to expose visibility, prompts, answers, citations, recommendations, audits, and selected writes.
Yes. Raw AI responses can be exported as CSV or JSON on all plans, alongside CSV data exports and PDF reports.
A native outbound event product is not publicly documented. The documented webhook is an inbound endpoint for sending crawler and referral logs into Agent Analytics.
Sources and related reading
See how AI talks about your brand
Enter your domain to get a free AI visibility report in under 60 seconds.