GEO/AEO

AEO (Answer Engine Optimization) gives your agent tools to understand and optimize for AI search. Track how LLMs like ChatGPT, Claude, Gemini, and Perplexity respond to queries, monitor brand mentions in AI responses, scrape ChatGPT Search results with citations, and analyze AI search volume trends.

CLI First

# LLM response check
naive aeo llm-responses chatgpt --keyword "best project management software 2026"

# LLM mentions tracking
naive aeo llm-mentions search --keyword "project management software"

# AI keyword volume
naive aeo ai-keywords --keywords "project management software"

SDK

import { Naive } from "@usenaive-sdk/server";
const naive = new Naive({ apiKey: process.env.NAIVE_API_KEY! });

// Your own account, or a tenant via naive.forUser(id).aeo
await naive.aeo.llmResponses("chatgpt", { user_prompt: "best project management software 2026", model_name: "gpt-4o-mini" });
await naive.aeo.aiKeywords({ keywords: ["project management software"] });
await naive.aeo.llmMentions("search", { /* params */ });

Per-user and AccountKit-gated; usage is metered against the tenant’s plan (see Customer Billing). Full surface in the aeo sub-client; also available to agents via agentTools().

Endpoints

API	Endpoint	Description	Speed	Cost
LLM Responses	`POST /v1/aeo/llm-responses/:platform`	Get responses from ChatGPT, Claude, Gemini, Perplexity	Live (~5-15s)	1.2 credits
LLM Responses	`POST /v1/aeo/llm-responses/:platform/task`	Async task submission (chatgpt, claude, gemini only)	Standard	1.2 credits
LLM Responses	`GET /v1/aeo/llm-responses/:platform/task/:id`	Retrieve async task results	GET	Free
LLM Responses	`GET /v1/aeo/llm-responses/:platform/tasks-ready`	Check for completed tasks	GET	Free
LLM Responses	`GET /v1/aeo/llm-responses/:platform/models`	List available models for a platform	GET	Free
LLM Scraper	`POST /v1/aeo/llm-scraper/chatgpt`	Scrape ChatGPT Search (advanced format)	Live	0.16 credits
LLM Scraper	`POST /v1/aeo/llm-scraper/chatgpt/html`	Scrape ChatGPT Search (HTML format)	Live	0.16 credits
LLM Scraper	`POST /v1/aeo/llm-scraper/chatgpt/task`	Async scraper task submission	Standard	0.08 credits
LLM Scraper	`GET /v1/aeo/llm-scraper/chatgpt/task/:id`	Retrieve scraper results (advanced)	GET	Free
LLM Scraper	`GET /v1/aeo/llm-scraper/chatgpt/task/:id/html`	Retrieve scraper results (HTML)	GET	Free
LLM Scraper	`GET /v1/aeo/llm-scraper/chatgpt/tasks-ready`	Check for completed scraper tasks	GET	Free
AI Keywords	`POST /v1/aeo/ai-keywords/search-volume`	AI search volume estimates	Live	0.4 + 0.004 per keyword
LLM Mentions	`POST /v1/aeo/llm-mentions/search`	Detailed mentions with counts and cited links	Live	2.4 + 0.024 per row
LLM Mentions	`POST /v1/aeo/llm-mentions/aggregated-metrics`	Consolidated metrics across dimensions	Live	2.4 + 0.024 per row
LLM Mentions	`POST /v1/aeo/llm-mentions/cross-aggregated-metrics`	Compare metrics across multiple targets	Live	2.4 + 0.024 per row
LLM Mentions	`POST /v1/aeo/llm-mentions/top-domains`	Most frequently mentioned domains	Live	2.4 + 0.024 per row
LLM Mentions	`POST /v1/aeo/llm-mentions/top-pages`	Top mentioned pages for a keyword	Live	2.4 + 0.024 per row
General	`POST /v1/aeo/id-list`	List task IDs across all sub-APIs	POST	Free

LLM Responses

Get responses from major AI models to see what they say about your brand, product, or topic.

LLM	Live Endpoint	Standard (async)	Models
ChatGPT	`POST /v1/aeo/llm-responses/chatgpt`	task, task/:id, tasks-ready	`GET .../chatgpt/models`
Claude	`POST /v1/aeo/llm-responses/claude`	task, task/:id, tasks-ready	`GET .../claude/models`
Gemini	`POST /v1/aeo/llm-responses/gemini`	task, task/:id, tasks-ready	`GET .../gemini/models`
Perplexity	`POST /v1/aeo/llm-responses/perplexity`	Live only — no async	`GET .../perplexity/models`

# Get ChatGPT's response to a query (live)
curl -X POST https://api.usenaive.ai/v1/aeo/llm-responses/chatgpt \
  -H "Authorization: Bearer nv_sk_your_key" \
  -H "Content-Type: application/json" \
  -d '{ "user_prompt": "best project management software 2026", "model_name": "gpt-4o-mini" }'

const response = await fetch("https://api.usenaive.ai/v1/aeo/llm-responses/chatgpt", {
  method: "POST",
  headers: {
    "Authorization": "Bearer nv_sk_your_key",
    "Content-Type": "application/json",
  },
  body: JSON.stringify({ user_prompt: "best project management software 2026", model_name: "gpt-4o-mini" }),
});
const data = await response.json();

Response:

{
  "results": [
    {
      "keyword": "best project management software 2026",
      "model": "gpt-4o",
      "response": "Here are some of the best project management tools in 2026:\n\n1. **Asana** — Great for team collaboration...\n2. **Linear** — Developer-focused with fast UI...\n3. **Notion** — All-in-one workspace...",
      "cited_domains": ["asana.com", "linear.app", "notion.so"],
      "cited_urls": [
        "https://asana.com/product",
        "https://linear.app/features"
      ]
    }
  ],
  "credits_used": 1.2
}

Parameters

Param	Type	Required	Default	Description
`user_prompt`	string	Yes	—	Query to send to the LLM
`model_name`	string	Yes	—	Model to use. Each platform accepts only its own models — list them via `GET .../{platform}/models`

The CLI maps --keyword to user_prompt. The model field is model_name, and each platform only accepts its own models (an OpenAI model name is rejected for Claude/Gemini/Perplexity). If --model is omitted, a valid per-platform default is used: chatgpt → gpt-4o-mini, claude → claude-sonnet-4-5, gemini → gemini-2.5-flash, perplexity → sonar. Perplexity is Live-only and supports only sonar, sonar-pro, sonar-reasoning-pro. Do not send location/language to LLM Responses.

When to use

Monitoring what LLMs say about your brand or product
Comparing how different AI models describe a topic
Identifying which competitors get mentioned in AI responses
Auditing AI search visibility across platforms

LLM Scraper

Scrape ChatGPT Search results including citations, source links, and structured response data. Two output formats: advanced (structured JSON) and HTML (raw rendered output).

Endpoint	Description	Method
`POST /v1/aeo/llm-scraper/chatgpt`	Scrape ChatGPT Search (advanced format)	Live
`POST /v1/aeo/llm-scraper/chatgpt/html`	Scrape ChatGPT Search (HTML format)	Live
`POST /v1/aeo/llm-scraper/chatgpt/task`	Submit async scraping task	Standard
`GET /v1/aeo/llm-scraper/chatgpt/task/:id`	Get results (advanced format)	GET
`GET /v1/aeo/llm-scraper/chatgpt/task/:id/html`	Get results (HTML format)	GET
`GET /v1/aeo/llm-scraper/chatgpt/tasks-ready`	Check for completed tasks	GET

Parameters

Param	Type	Required	Default	Description
`keyword`	string	Yes	—	Query to send to ChatGPT Search
`location_code`	number	No	—	Location code
`language_code`	string	No	—	Language code

When to use

Getting full ChatGPT Search results with source citations
Analyzing which sources ChatGPT cites for your keywords
Comparing live ChatGPT output vs. LLM Responses API data

AI Keyword Data

Get search volume estimates for keywords in AI search engines.

curl -X POST https://api.usenaive.ai/v1/aeo/ai-keywords/search-volume \
  -H "Authorization: Bearer nv_sk_your_key" \
  -H "Content-Type: application/json" \
  -d '{ "keywords": ["best ai tools", "chatbot for business", "ai agent framework"] }'

const response = await fetch("https://api.usenaive.ai/v1/aeo/ai-keywords/search-volume", {
  method: "POST",
  headers: {
    "Authorization": "Bearer nv_sk_your_key",
    "Content-Type": "application/json",
  },
  body: JSON.stringify({ keywords: ["best ai tools", "chatbot for business", "ai agent framework"] }),
});
const data = await response.json();

Response:

{
  "results": [
    { "keyword": "best ai tools", "ai_search_volume": 48200, "competition": "medium" },
    { "keyword": "chatbot for business", "ai_search_volume": 22100, "competition": "high" },
    { "keyword": "ai agent framework", "ai_search_volume": 8400, "competition": "low" }
  ],
  "credits_used": 0.8
}

Parameters

Param	Type	Required	Default	Description
`keywords`	string[]	Yes	—	Keywords to check (up to 1000)
`location_code`	number	No	2840	Location code (United States)
`language_code`	string	No	en	Language code

DataForSEO requires both a location and a language. location_code (2840) and language_code (en) are applied automatically when omitted; the CLI’s --location/--language flags override them.

When to use

Estimating AI search demand before creating content
Comparing traditional vs. AI search volume
Prioritizing keywords for AEO optimization

LLM Mentions

Track how often brands, domains, and keywords appear in LLM responses. All endpoints use Live method.

Endpoint	Description
`search`	Detailed mentions data with counts and cited links
`aggregated-metrics`	Consolidated metrics across dimensions
`cross-aggregated-metrics`	Compare metrics across multiple targets
`top-domains`	Most frequently mentioned domains for a keyword
`top-pages`	Top mentioned pages for a keyword

# Track brand mentions in LLM responses
curl -X POST https://api.usenaive.ai/v1/aeo/llm-mentions/search \
  -H "Authorization: Bearer nv_sk_your_key" \
  -H "Content-Type: application/json" \
  -d '{ "target": [{ "domain": "salesforce.com" }] }'

const response = await fetch("https://api.usenaive.ai/v1/aeo/llm-mentions/search", {
  method: "POST",
  headers: {
    "Authorization": "Bearer nv_sk_your_key",
    "Content-Type": "application/json",
  },
  body: JSON.stringify({ target: [{ domain: "salesforce.com" }] }),
});
const mentions = await response.json();

Response:

{
  "results": [
    {
      "keyword": "CRM software",
      "target": "salesforce.com",
      "mention_count": 47,
      "llm_platforms": ["chatgpt", "claude", "gemini", "perplexity"],
      "cited_urls": [
        "https://salesforce.com/products/crm",
        "https://salesforce.com/solutions/small-business"
      ],
      "context_snippets": [
        "Salesforce remains one of the most widely used CRM platforms..."
      ]
    }
  ],
  "credits_used": 4.8
}

Parameters

Param	Type	Required	Default	Description
`target`	object[]	Yes	—	Array of objects, each `{ "keyword": "..." }` or `{ "domain": "..." }`. A top-level `keyword`/`domain` string is invalid
`location_code`	number	No	2840	Location code
`language_code`	string	No	en	Language code
`date_from`	string	No	—	Start date (YYYY-MM-DD)
`date_to`	string	No	—	End date (YYYY-MM-DD)

The target array is the only accepted way to pass keywords/domains — DataForSEO rejects a top-level keyword or domain string with “POST Data Is Invalid”. The CLI builds it automatically: search/top-domains/top-pages from --keyword (→ target: [{ keyword }]), aggregated-metrics/cross-aggregated-metrics from --target <domain> (→ target: [{ domain }]).

When to use

Monitoring your brand’s presence in AI-generated answers
Benchmarking competitor visibility across LLMs
Identifying which domains dominate AI search for your keywords
Tracking mention trends over time

Error Handling

Error	Cause	Recovery
`insufficient_credits`	Not enough credits	Check balance with `GET /v1/status`
`invalid_llm`	Unknown LLM platform name	Use chatgpt, claude, gemini, or perplexity
`invalid_keywords`	Empty or malformed keyword list	Provide at least one keyword string
`invalid_input`	Platform doesn’t support async tasks	Perplexity is live-only; use chatgpt, claude, or gemini for tasks
`provider_error`	Data provider returned an error	Retry after a few seconds

Perplexity only supports the Live method — no async task endpoints. For batch processing across multiple LLMs, submit async tasks for ChatGPT, Claude, and Gemini, then use live calls for Perplexity.

Typical Workflow

Agent receives task: "Audit our brand visibility in AI search"
    │
    ├─ POST /v1/aeo/llm-responses/chatgpt             → Check ChatGPT's response
    │   { user_prompt: "best [product category]", model_name: "gpt-4o-mini" }
    │   → Full response with brand mentions, cited domains
    │
    ├─ POST /v1/aeo/llm-responses/claude               → Check Claude's response
    │   { user_prompt: "best [product category]", model_name: "claude-sonnet-4-5" }
    │   → Compare across models
    │
    ├─ POST /v1/aeo/llm-mentions/search                → Track mention frequency
    │   { target: [{ domain: "yourbrand.com" }] }
    │   → Mention count, cited URLs, context snippets
    │
    ├─ POST /v1/aeo/llm-mentions/top-domains           → See who dominates
    │   { target: [{ keyword: "product category" }] }
    │   → Ranked list of most-mentioned domains
    │
    └─ POST /v1/aeo/ai-keywords/search-volume          → Estimate AI search demand
        { keywords: ["product category", "brand name"] }
        → Volume estimates for AI search

Credit Costs

Operation	Credits
LLM Responses live	1.2
LLM Responses task_post	1.2
LLM Scraper live	0.16
LLM Scraper task_post	0.08
AI Keywords live	0.4 + 0.004 per keyword (0.8 for 100 keywords)
LLM Mentions live	2.4 + 0.024 per returned row (4.8 at the default 100 rows, 26.4 at the 1000-row maximum)
Models GET	Free
Task retrieval (task/:id, tasks-ready)	Free
id-list	Free

LLM Mentions and AI Keywords are metered, not flat: the provider bills us per returned row and per keyword, so the price is computed from the request’s own limit / keywords. limit accepts 1–1000 and defaults to 100. LLM Responses passes the model provider’s token cost through, so a long answer on an expensive model can settle above the list price — credits_used on the response is always the amount charged.

Getting Started

Agent Profiles

Architecture

Primitives

CLI First

SDK

Endpoints

LLM Responses

Parameters

When to use

LLM Scraper

Parameters

When to use

AI Keyword Data

Parameters

When to use

LLM Mentions

Parameters

When to use

Error Handling

Typical Workflow

Credit Costs

See also (blog)

​CLI First

​SDK

​Endpoints

​LLM Responses

​Parameters

​When to use

​LLM Scraper

​Parameters

​When to use

​AI Keyword Data

​Parameters

​When to use

​LLM Mentions

​Parameters

​When to use

​Error Handling

​Typical Workflow

​Credit Costs

​See also (blog)

CLI First

SDK

Endpoints

LLM Responses

Parameters

When to use

LLM Scraper

Parameters

When to use

AI Keyword Data

Parameters

When to use

LLM Mentions

Parameters

When to use

Error Handling

Typical Workflow

Credit Costs

See also (blog)