Aiso API beta

Aiso API

Programmatic access to your AI visibility data. Query tracked prompts, brand mentions, visibility, sentiment, position, and the sources cited by ChatGPT and Gemini. All endpoints return JSON and authenticate with a per-account API key.

This API is in beta. Endpoints and response shapes may evolve.

One key per account

Add x-aiso-project-id to scope a call.

Brand visibility

Mentions, sentiment, and position.

Source intelligence

Domains and URLs cited by AI.

Authentication

Include your API key in the x-api-key header on every request. You have one key per account. For project-specific calls, also send x-aiso-project-id with the project UUID. Generate a key in the Aiso app . Keys are stored as SHA-256 hashes, so the raw key is shown once on creation and cannot be retrieved again.

curl
curl -X GET "https://app.getaiso.com/api/v1/projects" \
  -H "x-api-key: aiso_abc123..."

curl -X GET "https://app.getaiso.com/api/v1/prompts" \
  -H "x-api-key: aiso_abc123..." \
  -H "x-aiso-project-id: YOUR_PROJECT_UUID"

Rate limits

Default limits are 10 requests per minute and 20 requests per day per API key. Admins can adjust these per user. Every response includes headers you can use to manage retries.

HeaderDescription
X-RateLimit-LimitMax requests in the current window.
X-RateLimit-RemainingRequests remaining before throttling.
X-RateLimit-ResetSeconds until the window resets.

Base URL

Base URL
https://app.getaiso.com/api/v1

Models and filtering

Aiso tracks two AI models. Use gpt for OpenAI ChatGPT with web search and gemini for Google Gemini with Google Search grounding. Report endpoints accept a flexible query body with dates, pagination, dimensions, filters, and sort order.

ParameterTypeDescription
start_datestring (YYYY-MM-DD)Snapshots on or after this date.
end_datestring (YYYY-MM-DD)Snapshots on or before this date.
limitnumber (1–10,000)Max rows. Default 1,000.
offsetnumberSkip N rows. Default 0.
dimensionsstring[]Break down by prompt_id, model_id, date, week, or month.
filtersobject[]{ field, operator: "in" | "not_in", values: [] }
order_byobject[]{ field, direction: "asc" | "desc" }
Report body
{
  "start_date": "2025-09-01",
  "end_date": "2025-10-01",
  "dimensions": ["date"],
  "filters": [
    { "field": "model_id", "operator": "in", "values": ["gpt"] }
  ],
  "order_by": [{ "field": "visibility", "direction": "desc" }],
  "limit": 100
}
GET/projects

Project

Returns project metadata. Without x-aiso-project-id, it lists all projects with summary fields. With the header, it returns one project including full domain_data.

Response (single project)
{
  "data": {
    "id": "uuid",
    "domain": "example.com",
    "company_name": "Example Inc",
    "niche": "CRM Software",
    "topics": ["pricing", "features", "alternatives"],
    "competitors": ["Competitor A", "Competitor B"],
    "geography": "US",
    "current_step": 2,
    "has_tracking": false,
    "created_at": "2025-09-22T...",
    "updated_at": "2025-10-01T...",
    "domain_data": {
      "domain": "example.com",
      "companyName": "Example Inc",
      "niche": "CRM Software",
      "topics": ["pricing", "features", "alternatives"],
      "competitors": ["Competitor A", "Competitor B"],
      "geography": "US",
      "language": "English",
      "entityType": "brand"
    }
  }
}
POST/projects

Create a project

Requires domain only. The server runs domain analysis once, saves domain_data, and returns it. Optional fields: country, language, entity_type, and prompts. The response uses the same shape as the single-project GET and may take up to about two minutes.

Request
{
  "domain": "example.com",
  "country": "US",
  "language": "English",
  "entity_type": "brand"
}
GET/prompts

Prompts

List the prompts tracked in the project. Optionally filter by topic with ?topic=pricing.

Response
{
  "data": [
    {
      "id": "prompt-uuid",
      "text": "What is the best CRM for small businesses?",
      "topic": "alternatives",
      "intent": "informational",
      "funnel_stage": "awareness",
      "branded": "non-branded",
      "source": "generated"
    }
  ],
  "total_count": 42
}
GET/brands

Brands

List the tracked brands, meaning your brand plus its competitors.

Response
{
  "data": [
    { "name": "Example Inc", "is_own": true, "is_competitor": false },
    { "name": "Competitor A", "is_own": false, "is_competitor": true }
  ],
  "total_count": 5
}
GET/snapshots

Snapshots

List analysis snapshots. Each snapshot is one visibility analysis run.

Response
{
  "data": [
    {
      "id": "snap-uuid",
      "status": "complete",
      "prompts_count": 42,
      "run_at": "2025-10-01T12:00:00Z",
      "completed_at": "2025-10-01T12:05:00Z"
    }
  ],
  "total_count": 8
}
POST/reports/brands

Brand mentions

Measure visibility for your brand and competitors across tracked prompts. Each result includes mention rate, sentiment, and average position.

Filters: model_id, prompt_id, topic_id, and brand_id. Dimensions: prompt_id, model_id, date, week, month.

Request
curl -X POST "https://app.getaiso.com/api/v1/reports/brands" \
  -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
  "start_date": "2025-09-01",
  "end_date": "2025-10-01",
  "dimensions": ["date"],
  "filters": [
    { "field": "model_id", "operator": "in", "values": ["gpt"] }
  ],
  "order_by": [{ "field": "visibility", "direction": "desc" }],
  "limit": 100
}'
Response
{
  "data": [
    {
      "brand": { "name": "Example Inc" },
      "mention_count": 28,
      "visibility": 0.67,
      "visibility_count": 28,
      "visibility_total": 42,
      "sentiment": 72,
      "sentiment_sum": 14,
      "sentiment_count": 28,
      "position": 2.3,
      "position_sum": 64.4,
      "position_count": 28,
      "date": "2025-09-15"
    }
  ],
  "total_count": 150
}
FieldDescription
visibilityMention rate from 0 to 1. How often the brand appears across tracked prompts.
mention_countPrompt responses where the brand was mentioned.
sentimentNormalized 0 to 100 score. 50 is neutral, above 50 is positive, below 50 is negative.
positionAverage rank position when mentioned, where 1 is first. Null if not mentioned.
POST/reports/domains

Sources

Find the domains cited by AI models in their responses. Filter by model_id or domain, then order by used_count or citation_count.

Request
curl -X POST "https://app.getaiso.com/api/v1/reports/domains" \
  -H "x-api-key: YOUR_API_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "start_date": "2025-09-01",
    "order_by": [{ "field": "used_count", "direction": "desc" }],
    "limit": 20
  }'
Response
{
  "data": [
    {
      "domain": "g2.com",
      "domain_type": "ugc",
      "used_count": 34,
      "citation_count": 52,
      "citation_rate": 1.53
    }
  ],
  "total_count": 87
}
FieldDescription
used_countTimes URLs from this domain appeared in AI responses.
citation_countTotal inline citations from this domain.
citation_rateCitations per retrieval (citation_count / used_count).
domain_typeyou, competitor, ugc, editorial, corporate, reference, institutional, or other.
POST/reports/urls

Source URLs

Drill into individual cited pages. Filter with model_id, domain, or url.

Response
{
  "data": [
    {
      "url": "https://g2.com/products/example/reviews",
      "domain": "g2.com",
      "domain_type": "ugc",
      "title": "Example Reviews 2025 | G2",
      "used_count": 12,
      "citation_count": 18,
      "citation_rate": 1.5
    }
  ],
  "total_count": 245
}

Common recipes

Ready-made request bodies for the questions teams ask most.

Overall visibility for all brands

POST /reports/brands
{ }

Daily visibility trend for your brand

POST /reports/brands
{
  "dimensions": ["date"],
  "filters": [
    { "field": "brand_id", "operator": "in", "values": ["Your Brand"] }
  ]
}

Visibility by AI model (GPT vs Gemini)

POST /reports/brands
{
  "dimensions": ["model_id"]
}

Weekly visibility trend

POST /reports/brands
{
  "dimensions": ["week"],
  "start_date": "2025-08-01"
}

Top cited domains (GPT only)

POST /reports/domains
{
  "filters": [
    { "field": "model_id", "operator": "in", "values": ["gpt"] }
  ],
  "order_by": [{ "field": "used_count", "direction": "desc" }],
  "limit": 10
}

URL drilldown for a specific domain

POST /reports/urls
{
  "filters": [
    { "field": "domain", "operator": "in", "values": ["g2.com"] }
  ],
  "order_by": [{ "field": "citation_count", "direction": "desc" }]
}

MCP (Cursor & Claude)

Connect Cursor, Claude Desktop, or any MCP host directly to your Aiso account. The hosted MCP server wraps the v1 API as tools — same API key, no local install.

1. Create an API key

Go to API keys and create a key. Copy it when shown — it cannot be retrieved again.

2. Add to Cursor

Create .cursor/mcp.json in your project (or edit ~/.cursor/mcp.json globally), then restart Cursor or reload MCP from Settings.

mcp.json
{
  "mcpServers": {
    "aiso": {
      "url": "https://app.getaiso.com/api/v1/mcp",
      "headers": {
        "x-api-key": "aiso_your_key_here",
        "x-aiso-project-id": "your_project_uuid"
      }
    }
  }
}

For local dev, use http://localhost:3000/api/v1/mcp instead. Use list_projects to discover project IDs, then set x-aiso-project-id in the headers.

Legacy stdio clients

If your MCP host does not support remote URLs yet, proxy the hosted endpoint with mcp-remote:

mcp.json (stdio)
{
  "mcpServers": {
    "aiso": {
      "command": "npx",
      "args": [
        "-y",
        "mcp-remote",
        "https://app.getaiso.com/api/v1/mcp",
        "--header",
        "x-api-key:aiso_your_key_here",
        "--header",
        "x-aiso-project-id:your_project_uuid"
      ]
    }
  }
}

Project header

GET /projects without a project header lists all your projects. With x-aiso-project-id, it returns one project. All other project-scoped routes require the header, or ?project_id= on GET.

Setup and prompt writes

POST /projects accepts a domain and runs analysis once, so there is no separate analyze endpoint. Setup and prompt management are open to all API key holders. Each project allows one setup write and one prompt write via the API or MCP: further setup updates return 409 once setup exists, and further prompt writes return 409 once prompts exist. The initial POST /projects may include prompts in the same request. Prompt suggestion and generation require stored setup (domain_data) and do not write to the database, so call add_prompts to save.

Visibility reports

Snapshots, brand, domain, and URL reports, and fanout metrics require a completed visibility analysis run for the project. Until then, those tools return 403 with guidance to reach out at getaiso.com for help.

Available MCP tools

ToolDescription
get_accessAccount limits and optional project info
list_projectsAll projects for your account
get_projectOne project with full domain_data
list_promptsTracked prompts
list_brandsBrands and competitors
list_snapshotsAnalysis runs
brand_visibility_reportVisibility report
domain_reportDomain citations
url_reportURL citations
get_fanout_metricsQuery fanout analytics
get_fanouts_contextFanouts dashboard context
create_projectCreate project (domain analysis included)
update_projectUpdate project setup
add_promptsAdd prompts (one write per project)
update_promptsUpdate prompts (before first add_prompts only)
manage_topicCreate, update, or delete topics (before first add_prompts only)
suggest_topicsAI-suggest topics (requires setup, preview only)
suggest_promptsAI-suggest prompts (requires setup, preview only)
generate_topic_promptsGenerate prompts for a topic (requires setup, preview only)

Troubleshooting

StatusMeaning
400Missing x-aiso-project-id on a project-scoped route, or missing required body field
401Missing or invalid API key
403Visibility data not available yet (analysis not run)
409Setup or prompt write already completed for this project
429Rate limit exceeded

MCP endpoint: https://app.getaiso.com/api/v1/mcp · Admin usage: /admin/api

Need conversations data?

The Aiso API is for tracking your brand’s visibility and the sources used in AI results. For public access to Aiso’s conversations dataset, use our separate Apify API.

Open Conversations API

Errors

Error responses always return { "error": "message" }.

StatusMeaning
401Missing or invalid API key.
403Visibility data not available yet.
409Setup or prompt write already completed, or project limit reached.
404Project or resource not found.
429Rate limit exceeded.
500Internal server error.