Skip to content

Guides

DataForSEO alternatives with share of voice and sentiment in 2026

DataForSEO's AI Optimization API returns answers and brand mentions at $1.20 per 1,000, with no share of voice or sentiment. Eight APIs compute both, from $3.30.

A funnel filled with white cubes pours into a solid blue cylindrical container

SEO consultant, and more recently GEO, at MentionBox

DataForSEO's AI Optimization API hands you the answer, the links under it and the brands named inside it, at $1.20 per 1,000 ChatGPT or Gemini answers on its standard queue. It computes no share of voice, no brand position and no sentiment, and its LLM Mentions endpoints return a mentions count and an AI search volume rather than either of the two metrics a client asks about. Eight of the APIs we track return share of voice and sentiment per brand with a rate per 1,000 answers you can read before you sign, from $3.30 with SE Visible to $81.94 with AthenaHQ. Our pick of the group, after hands-on testing by Mathias C., is API for GEO at $13 per 1,000 answers with both metrics in the same response.

This guide is for the agency or brand team already collecting AI answers through DataForSEO, with a client now asking for a share of voice chart and a sentiment trend. It covers what the DataForSEO payload does contain, what the two missing metrics cost to compute yourself, the eight alternatives one by one, and the reason two vendors report two different sentiment scores for the same brand. MentionBox, which publishes this site, has a commercial link with API for GEO.

The alternatives at a glance

APIEngines, of the 10 we trackShare of voiceSentimentPrice per 1,000 answersEntry ticket that carries the keyResale
API for GEO10yesyes$13, down to $7 at volumepay as you go, no minimum publishedallowed
LLM Pulse10yesyes€30.67€299 a month, Scaleallowed
AthenaHQ10yesyes$81.94, plus an add-on priced on request$295 a month, Starterallowed
Scrunch9yesyesnot computable$500 a month, Agency Corenot disclosed
OtterlyAI7yesyes$15.75$189 a month, Standardnot disclosed
GEOly7yesyes$13.32$999 a month, Plusnot disclosed
Ahrefs Brand Radar8yesyes$20$129 a month, Lite, plus a $50 check packnot allowed
SE Visible4yesyes$3.30$99 a month, Basicnot allowed
Superlines9yesyes€14.04€379 a month, Growthnot allowed
SE Ranking Data API5yesno€36€45 wallet, 250,000 creditsnot allowed
DataForSEO AI Optimization, for reference5nono$1.20$50 minimum paymentno clause published

Published by each vendor on its pricing page and in its API reference, checked on October 23, 2026. Rates per 1,000 answers are our arithmetic on the published allowance of the cheapest plan that carries an API key. For API for GEO, four of the ten engines, Gemini, Google AI Overviews, DeepSeek and Meta AI, plus the webhooks, the OpenAPI spec and the official SDKs, were confirmed in our hands-on test on October 26, 2026 and are not yet documented on apiforgeo.com.

Three columns decide this switch, and the headline rate is rarely the one that does:

  • Whether the key sits on usage or on a subscription. DataForSEO and API for GEO bill what you send with no monthly floor. The other nine put the key on a plan you pay in full whether you call it twice or ten thousand times.
  • Whose prompts produce the numbers. SE Ranking's Data API and the Ahrefs prompt indexes answer from a stored database refreshed on their own schedule. The rest run your own prompt list.
  • What the terms allow. Five of the ten forbid resale or redistribution outright, three publish no clause, and three allow a client-facing product under your own brand.

What DataForSEO's AI Optimization API returns, and where it stops

The AI Optimization API is four endpoint families. The LLM Scraper "provides scraped results from popular LLM platforms". LLM Responses returns "structured responses from leading LLMs, including ChatGPT, Claude, Gemini, and Perplexity" through their official APIs. AI Keyword Data gives "search volume estimates and user intent insights based on keyword usage in AI tools". LLM Mentions gives "data on keyword, brand and website mentions in LLMs, including metrics like AI search volume, impressions and mentions count".

The scraper is the family that costs $1.20 per 1,000, and its ChatGPT advanced response is well stocked for a raw payload:

Field in the ChatGPT LLM Scraper responseWhat it carries
markdown"content of the element in markdown format"; live/html returns the raw HTML
sources"the sources the model actually cited or relied on in its final answer", with url, domain and title
search_resultsevery web search output the model retrieved
brand_entities"contains information on brands mentioned in the response"
fan_out_queries"related search queries derived from the main query"
chat_gpt_productstitle, price, currency, merchants, product ids, plus local businesses and ads

Field names and definitions from the DataForSEO API reference, checked on October 23, 2026. brand_entities and fan_out_queries are documented on the ChatGPT scraper and absent from the Gemini scraper reference.

curl -X POST https://api.dataforseo.com/v3/ai_optimization/chat_gpt/llm_scraper/task_post \
  -u "$DFS_LOGIN:$DFS_PASSWORD" \
  -H "Content-Type: application/json" \
  -d '[{
    "user_prompt": "best crm for a 20-person agency",
    "location_code": 2840,
    "language_code": "en",
    "web_search": true
  }]'

One answer comes back with its brands listed, and no number about those brands. There is no rank field for a brand, only the rank_group and rank_absolute of the answer elements themselves, and no sentiment field anywhere in the family: the LLM Mentions page offers "a piece of a solution" for sentiment analysis rather than a sentiment endpoint.

LLM Mentions is the endpoint most often mistaken for a substitute. Its live target metrics endpoint takes up to ten domains or keywords and returns, per location, language, platform and source domain, a mentions count and an ai_search_volume, read from a stored database of 420,701,746 prompts whose ChatGPT rows are limited to the United States in English. It bills $0.1 per task plus $0.001 per row. The figure describes a market rather than your tracked prompt list.

On integration it beats every alternative here: SDKs in Python, C#, TypeScript and Java, postback and pingback webhooks, up to 100 tasks per POST, up to 10,000 calls a minute, an OpenAPI repository and an MCP server. Our review of the DataForSEO AI Optimization API scores it 7.7 out of 10, with 10 on pricing and integration against 5.6 on the data returned, and the duel API for GEO vs DataForSEO sets the two payloads side by side.

What the two missing metrics cost to compute yourself

Share of voice is arithmetic: over a fixed set of collected answers, your brand's mentions divided by every tracked brand's mentions. Position is the rank of your first mention among the brands named in an answer, averaged over the answers that name you. Our explainer on GEO metrics gives the published formula of each one and the denominators that differ between vendors, and our walkthrough of tracking a brand in ChatGPT with an API writes them out as code.

The work sits in the entity resolution before the division:

  1. Collapse aliases onto one brand. Legal suffixes, product names, abbreviations, misspellings and translated names all count as the same mention, and the list grows every month.
  2. Decide what a mention is. A brand in the answer text, a brand only in a cited page title and a brand in a shopping card are three different events, and mixing them moves a share of voice by several points.
  3. Freeze the denominator and store the raw answers. Prompts, engines, locale and run count are the denominator of every metric here, and stored answers turn a change of counting rule into a recount instead of a month of fresh collection.

Sentiment is the expensive half, since it needs a classifier over every answer rather than a formula. DataForSEO's own LLM Responses endpoints can run that pass at $0.0002 per task on the standard queue plus the token cost charged by the model provider. The worked example in its documentation for ChatGPT gpt-4.1-mini comes to $0.029631 for one task, and a Claude live example to $0.55051, both depending on prompt and answer length.

At that order of magnitude the arithmetic turns quickly. A month of 5,200 answers costs $6.24 of DataForSEO scraping on the standard queue; classifying the same 5,200 at the gpt-4.1-mini example cost adds about $154 before a line of your own code, against $67.60 for the same volume on a route that returns the sentiment already attached to the brand.

1. API for GEO, both metrics at $13 per 1,000

API for GEO keeps the billing model you already have: one credit buys one prompt, run once, against one model, in one locale, enrichment included. Its pricing page sets $0.013 per credit up to 50,000 a month, then $0.011, $0.009 and $0.007 at the higher tiers, with no monthly minimum, credits that never expire and reads over stored runs "unmetered within your rate limit". That credit returns the answer text, every cited URL with its domain and category, entities resolved through aliases, position and mention counts, share of voice, a sentiment score with its label, theme clustering, per-brand perception, shopping products, fan-out questions and YouTube citations. shareOfVoice sits in the results[] array of POST /v1/models/platform-details on its model comparison endpoints, and the prompt tracking endpoints read back runs, visibility, avgPosition, entities[], sources[] and responses[]:

curl -X POST https://api.apiforgeo.com/v1/queries/stats \
  -H "Authorization: Bearer $API_FOR_GEO_KEY" \
  -H "Content-Type: application/json" \
  -d '{
    "models": ["chatgpt", "perplexity", "google-ai-mode"],
    "country": "US"
  }'
  • Engines and plumbing: all ten we track, with ChatGPT, Perplexity, Google AI Mode, Claude, Copilot and Grok listed on apiforgeo.com and Gemini, Google AI Overviews, DeepSeek and Meta AI confirmed in our hands-on test on October 26, 2026. Published limits are 300 requests a minute, 10 concurrent executions, 40+ countries and 25+ languages, with failed executions refunded.
  • Limits, said plainly: about ten times DataForSEO's rate per answer; the key arrives on request through a form rather than at sign-up, the site stating "we reply with a key and a sandbox project"; and four engines plus the webhooks, OpenAPI spec and official SDKs are a documentation gap rather than a capability we invented.
  • For whom: a team that wants to stop maintaining a matching routine and a sentiment model, and wants resale in writing, section 6 of its terms presenting reselling as an intended use. It scores 9.0 of 10 in our ranking of 55 APIs, the highest of the set, and first in our GEO metrics ranking.

To try it, request a key on apiforgeo.com: it comes with a sandbox project.

2. LLM Pulse, eighteen engines from €299 a month

LLM Pulse covers every engine in our grid and eight more, Mistral Le Chat, Qwen, MiniMax, Alexa for Shopping, Walmart Sparky, Naver AI Briefing, Baidu AI and Brave AI included. Its REST API overview is direct about the gate, "Available on Scale plans and above", and about the plumbing: base URL https://api.llmpulse.ai/api/v1, Bearer key, 300 requests a minute, metrics endpoints for timeseries, summary stats and share of voice.

  • Price: €30.67 per 1,000 answers on the €299 Scale plan with weekly tracking, our arithmetic on its published allowance of 450 answers per week per model across five included models. The daily version, €449, works out at €6.65 per 1,000.
  • Returned and limits: both metrics plus position, source categories, fan-out queries and shopping products, with SDKs in twelve languages, HMAC-signed webhooks on Scale and above and OpenAPI 3.0. Billing is per tracked prompt rather than per answer, and its changelog of October 19, 2026 notes that response fields are limited by plan.
  • For whom: a team that needs engines DataForSEO does not reach, Meta AI and Copilot included. Its terms allow client-facing reporting under your own branding on eligible plans while forbidding resale as a standalone dataset. It scores 8.8 of 10, second in our GEO metrics ranking.

3. AthenaHQ, ten engines with the API as a paid add-on

AthenaHQ publishes the most explicit share of voice of the group. Its cumulative share of voice endpoint is a POST /api/v1/metrics/share-of-voice/cumulative returning per brand a name, a value from 0 to 100, an absolute_value mention count, an is_self flag and a relative_mention_rate defined as "the entry's mentions as a percentage of responses that mention at least one tracked brand". The percentage and the raw count in the same row is what lets you re-derive the figure under your own denominator.

  • Price: $81.94 per 1,000 answers, from $295 a month for 3,600 credits on Starter, where "1 credit = 1 AI response". API access on Starter is a paid add-on whose price is not published, activation "fetches live, prorated pricing data from the billing system", so the real cost per answer is higher and unknown.
  • Returned and limits: both metrics plus position, source categories, shopping metrics and attribute observations on all ten engines plus Mistral, with 60 requests a minute, an OpenAPI spec and embeddable dashboards. The collection method is not published, webhooks are not documented, and the free Essential plan carries no API.
  • For whom: a brand team that wants every engine and a share of voice it can audit, and will ask for the add-on price in the same email as the demo. It scores 7.4 of 10, third in our GEO metrics ranking, and its agency page offers discounted licences bundled with your own managed services.

4. Scrunch, both metrics from $500 a month

Scrunch returns the full computed set on nine of the ten engines we track, Meta AI included. Its authentication page states that "API key creation is available on the Agency and Enterprise plans", and its help centre adds that single-brand plans include no API access, with the raw Responses API reserved for Enterprise.

  • Price: $500 a month for Agency Core, the cheapest plan with a key. No prompt or answer volume is published for it, so no rate per 1,000 can be computed; billing is stated as the number of AI responses collected rather than API call volume.
  • Returned and limits: both metrics plus position, source categories, fan-out queries and shopping products, with POST /{brand_id}/collections triggering an on-demand collection and retrieval that "does not count against your prompt quota". Webhooks are marked not generally available, rate limits are not disclosed, and data appears 20 to 60 minutes after a brand is created for small organisations, up to 12 hours for large ones.
  • For whom: an agency that wants both metrics on nearly every engine and prefers a flat fee to a meter. Its terms cover neither resale nor white label while its agency page offers "an agency-branded AI search product", so get that clause into your order form. It scores 7.0 of 10, fourth in our GEO metrics ranking.

5. OtterlyAI, a net sentiment score from $189 a month

OtterlyAI has the most report-shaped response here. One call to GET /v1/reports/brand/{reportId}/stats, with a required date range and country, returns summary.shareOfVoice and summary.averagePosition, a brandPositionHistory, and a competitorBrandsAnalysis.brandMentions[] array where each brand carries its own shareOfVoice, averagePosition and a sentiment object split into positive, neutral, negative and nss, a net sentiment score.

  • Price: $15.75 per 1,000, from $189 a month for 100 prompts on four engines tracked daily, which is 12,000 answers over a 30-day month. Extra prompts cost $99 per 100 a month, the AI Mode and Gemini pack $59, Claude $109.
  • Returned and limits: 7 of 10 engines, missing Grok, DeepSeek and Meta AI. The $29 Lite plan carries no API, the API request allowance is a second quota at 2,000 a month on Standard, and the full terms sit in a PDF whose text we could not extract, so we treat the resale clause as not disclosed. OtterlyAI states it "does not currently offer a native white-label option" while offering Looker Studio reports with agency branding.
  • For whom: a brand team that wants a net sentiment score and a share of voice in one call, read into a monthly report without a developer week. It scores 6.7 of 10, fifth in our GEO metrics ranking.

6. GEOly, one synchronous endpoint from $999 a month

GEOly exposes its whole product as one call. Its Data API describes "one tool, one endpoint, one synchronous JSON response": a POST https://app.geoly.ai/api/v1/tools/{tool}, with GET /api/v1/tools listing what exists, calls priced at 1 credit for a light tool, 3 for a standard one and 10 for a deep one, and rejected calls not billed.

  • Price: $13.32 per 1,000 answers, from $999 a month on Plus and its published 75,000 responses a month for 500 prompts. The $49 Basic plan carries no API.
  • Returned and limits: share of model and share of mention, sentiment in five classes with an average score, a rank view, source ownership split between brand-owned domains, third-party platforms and competitors, fan-out queries and shopping products, on 7 of 10 engines with no Claude, DeepSeek or Meta AI. The collection method is not published, API calls consume MCP credits as a second meter, and the terms publish no resale clause while forbidding use to "build or train a competing service".
  • For whom: a team already paying for a platform at that level and wanting its figures in a warehouse without a second vendor. It scores 6.6 of 10, sixth in our GEO metrics ranking.

7. Ahrefs Brand Radar, sentiment by aspect at $20 per 1,000

Ahrefs Brand Radar is the route most SEO teams already have a login for, and its Get AI responses endpoint returns a field shape the others do not: brand_sentiment, one entry per mention of a tracked brand, carrying the brand, the sentiment as positive, neutral, negative or mixed, an aspect, and the snippet that justified it. Share of voice has its own /sov-overview and /sov-history endpoints.

  • Price: $20 per 1,000 answers on the custom prompts path, where one check is "1 prompt execution × 1 LLM × 1 location" and the Basic pack is $50 for 2,500 checks, falling to $0.015 per check on the $100 Growth pack and $0.010 on the $250 Scale pack. The $129 Lite subscription sits on top with 150 checks included, and reading custom prompt data back through the API is free and "does not consume any API units".
  • Returned and limits: 8 of 10 engines, missing DeepSeek and Meta AI, with Claude costing 8 checks per update, which puts Claude answers at $160 per 1,000 on the Basic pack. No brand position field is documented; the Ahrefs prompt database needs the Brand Radar AI add-on from $199 a month per index and its chatbot rows refresh "once a month"; and clause 12.4 of the terms licenses the data "solely for the purposes of your internal business operations".
  • For whom: an in-house SEO team with an Ahrefs seat, tracking its own prompt list and reporting internally. It scores 6.0 of 10.

8. SE Visible, the cheapest computed layer at $3.30 per 1,000

SE Visible is SE Ranking's dedicated AI visibility product and the cheapest route here to a share of voice and a sentiment label. Its API reference exposes projects, locations, engines, brands, topics, prompts, sources and subscription on https://api.seranking.com/v1/se-visible, with usage "governed by your plan limits, not by per-request credits", and source filters that split content into listicle, alternative, comparison, review, how-to guide, FAQ page and article.

  • Price: $3.30 per 1,000 answers, from $99 a month on Basic and its published figure of roughly 30,000 AI answers analysed a month for 200 prompts and 3 projects. Core is $189 for 450 prompts, Plus $355 for 1,000.
  • Returned and limits: response text, cited URLs with their type, brand mentions, average position, share of voice and sentiment as positive, neutral or negative, refreshed daily. Only 4 of the 10 engines we track, with the FAQ saying "Claude will be added soon"; the API is marked beta and the pages do not say which plans include it; and SE Ranking's terms limit use to the customer's internal use.
  • For whom: an in-house team whose report covers Google and ChatGPT only, on a budget that stops short of $300 a month. It scores 5.8 of 10, ninth in our GEO metrics ranking.

Three more that compute both, with a catch

  • Superlines returns both metrics on nine engines through a single entry point, https://api.superlines.io with an endpoint parameter, includeSentiment and includeShareOfVoice defaulting to true. The key starts on the €379 Growth plan, €14.04 per 1,000 answers for 300 prompts on three engines collected daily. Its documentation is the only one here to name the method engine by engine: interface for ChatGPT, AI Overviews, AI Mode, Perplexity, Gemini, Copilot and Grok, official API for Claude, DeepSeek and Mistral. Section 2.5 of its terms forbids distributing, selling or sublicensing the technology. It scores 5.6 of 10.
  • AI Peekaboo returns share of voice, position and sentiment per brand on seven engines from $30 a month, with the plan configured as brands times prompts times frequency times models, so no rate per 1,000 answers can be computed from the published cards. It scores 4.6 of 10.
  • SE Ranking's Data API publishes a share of voice and no sentiment, and its answers come from SE Ranking's own prompt database rather than your prompt list, "refreshed on a monthly basis". Its AI Search endpoints cost 800 credits for the by-engine time series, 7,500 flat for the leaderboard that carries share_of_voice and 200 per returned prompt, against €45 for 250,000 credits, which is €36 per 1,000 answers returned. It scores 4.3 of 10.

Both metrics, no published price

Four names return share of voice and sentiment with no rate you can plan against. Profound captures from nine engines in the browser and documents visibility_score, share_of_voice, average_position and sentiment_claims by asset and theme, with the REST API on the Enterprise plan only, a case our guide to Profound alternatives with a self-serve API covers in full. Peec AI publishes its formulas, which is rare and useful, and limits API access to Enterprise billed annually on request. Conductor has a dedicated sentiment endpoint, POST /data-api/v1/async/ai_search_sentiments, where "each row represents a single sentiment mention for a brand or product within an AI response", on eight engines with no share of voice. seoClarity ArcAI covers nine engines with both metrics and publishes no price. All four forbid resale except Conductor, whose terms do not address it.

If you would rather keep a scraper and buy only part of the analysis, cloro costs $4 per 1,000 ChatGPT answers on its $30 Lite plan, more than three times DataForSEO, and buys a collection method documented at endpoint level plus 250+ countries with US state targeting; it computes neither metric and its terms forbid reselling the service. Two APIs return sentiment without a share of voice, which is worth knowing because the division is the easy half: Promptwatch at $15.83 per 1,000 on a $95 plan, with sentimentScore, toneOfVoice and a brandMentions array carrying position and sentiment per brand, resale allowed; and Rankability at $2.94 per 1,000 on a $97 plan covering all ten engines, with resale forbidden in writing.

Sentiment comes back in four different shapes

Share of voice at least has a recognisable formula, and our GEO metrics explainer covers the denominators that differ between vendors. Sentiment is less settled, and the shape of the field decides what you can chart.

APIHow sentiment is returned
API for GEOa score with its label, plus theme clustering and per-brand perception
OtterlyAIcounts of positive, neutral and negative plus nss, a net sentiment score
Peec AIone 0 to 100 score, published as ((sum(sentiment_sum) / sum(sentiment_count)) / 2 + 0.5) * 100
Ahrefs Brand Radarone row per mention: brand, positive, neutral, negative or mixed, an aspect, and the snippet
Conductorone row per mention, with a sentiment_category among nine, a score, a snippet and its source URL
GEOlyfive classes, positive, neutral, mixed, negative and unknown, plus an average score
Promptwatchsentiment as POSITIVE, NEGATIVE or NEUTRAL, plus sentimentScore and toneOfVoice
SE Visiblepositive, neutral or negative

Field names and scales from each vendor's API reference, checked on October 23, 2026.

Two consequences for whoever builds the chart. A single score and a distribution are not interchangeable: a brand with 40% positive and 20% negative mentions and a brand with 60% neutral mentions can land on the same average, and only the distribution shows the difference. And per-mention rows let you aggregate your own way, so Ahrefs and Conductor allow sentiment on price to be reported separately from sentiment on service, where a normalised score cannot be split back apart.

What a 5,200-answer month costs on each route

Take ten clients, 30 prompts each, four engines, collected weekly: 10 × 30 × 4 × 52 ÷ 12, about 5,200 answers a month.

APIEntry ticket that carries the keyAnswers it collects at that priceBill for a 5,200-answer month
API for GEOnone, pay as you goany volume, credits never expire$67.60
SE Visible$99 a month, Basicabout 30,000, 200 prompts$99.00
Ahrefs Brand Radar$129 a month, Lite, plus check packs150 checks included, then $50 per 2,500$129 plus $101 of checks
OtterlyAI$189 a month, Standard12,000, but 100 prompts$189 plus $198 for 200 extra prompts
LLM Pulse€299 a month, Scale weeklyabout 9,750€299.00
AthenaHQ$295 a month, Starter3,600 creditsover $295, API add-on not published
Superlines€379 a month, Growthabout 27,000, 300 prompts on 3 engines€379.00
Scrunch$500 a month, Agency Corenot published$500.00
GEOly$999 a month, Plus75,000, 500 prompts$999.00
SE Ranking Data API€45 wallet, 250,000 credits1,250 prompts returned€187.20
DataForSEO, no metrics$50 minimum paymentabout 41,667 answers of credit$6.24 of usage

Our arithmetic on the rates and allowances each vendor published, checked on October 23, 2026, under three stated assumptions: a 30-day month, every prompt running on every included engine at the plan's own frequency, and one country per prompt. The subscriptions collect more than 5,200 answers at their own frequency and bill their floor either way.

The pay-as-you-go routes scale down: at 1,000 answers a month API for GEO costs $13 and DataForSEO $1.20, where every subscription still costs its floor, since a fraction of a plan is not for sale. The subscriptions scale up better: at 30,000 answers a month SE Visible is $99 against $390 for the same volume on API for GEO, and LLM Pulse's daily Scale plan at €449 for about 67,500 answers beats everything here per answer.

Mistakes we see on this switch

  1. Treating a mention count as a share of voice. DataForSEO's LLM Mentions gives mentions and AI search volume over a stored prompt database. A share of voice is a ratio over your own tracked prompts, and the two move independently.
  2. Buying the sentiment and keeping the matching. A vendor-computed share of voice runs on the brand list you upload, so a stale alias list produces a wrong number in a new way.
  3. Comparing an entry price to a rate per answer. $99 a month and $13 per 1,000 answers compare only once you state a volume, so write your monthly answer count down before you read any pricing page.
  4. Assuming the API sits on the plan you trialled. Scrunch puts key creation on Agency and Enterprise, LLM Pulse on Scale, Superlines on Growth, OtterlyAI on Standard, GEOly on Plus, and AthenaHQ sells it as an add-on priced on request.
  5. Skipping the resale clause. Ahrefs, SE Visible, SE Ranking, Superlines, Rankability, cloro and Peec AI forbid resale or redistribution in writing, and for a client-facing product that line outranks every other column.
  6. Switching vendor without noting the break. Two vendors asking the same engine the same question on the same day get two different answers, and their sentiment scales do not line up. One line under the chart with the date of the change saves the meeting where a drop looks like a crisis.

Our view

  1. You want both metrics, per-answer billing, and the right to bill clients. API for GEO is our pick: share of voice, position, sentiment, themes and per-brand perception on all ten engines we track, at $13 per 1,000 answers down to $7, no monthly floor, resale as an intended use. The honest trade-offs, in our view, are the rate, about ten times DataForSEO's per answer, and the documentation gap our test of October 26, 2026 had to fill.
  2. Your report covers Google and ChatGPT, and the budget is tight. SE Visible at $99 a month and $3.30 per 1,000 is the cheapest computed layer we found. Four engines, a beta API and a clause limiting use to your own organisation are the price of that.
  3. You need engines DataForSEO does not reach. LLM Pulse covers all ten plus eight more, with twelve SDKs, signed webhooks and €6.65 per 1,000 on its daily Scale plan, the best full-coverage rate here above roughly 20,000 answers a month.

Our GEO metrics ranking scores every API that returns share of voice, position and sentiment, and the agency and enterprise brand rankings reweight the same criteria by what each profile pays for. Whichever you pick, run your own 50 prompts through it alongside DataForSEO for two weeks, then compare the brand lists before you compare the metrics: if the two disagree about who was mentioned, the share of voice disagreement is downstream of that.

Frequently asked questions

Does DataForSEO return share of voice or sentiment?

No. Its AI Optimization API returns the answer in markdown, the sources the model cited with their url, domain and title, a `brand_entities` item that "contains information on brands mentioned in the response", the fan-out queries and the ChatGPT shopping products. No field carries a share of voice, a brand position or a sentiment label, and the LLM Mentions page presents itself as "a piece of a solution" for sentiment analysis rather than a sentiment endpoint. Checked on October 23, 2026.

Which DataForSEO alternative is cheapest for share of voice and sentiment?

SE Visible, SE Ranking's AI visibility product, at $3.30 per 1,000 AI answers on its $99 Basic plan, which publishes roughly 30,000 answers analysed a month for 200 prompts. It returns share of voice, average position and positive, neutral or negative sentiment, and it covers four of the ten engines we track: ChatGPT, Gemini, Google AI Mode and Google AI Overviews. Its API is marked beta, and SE Ranking's terms limit use to the customer's internal use, so an agency reselling the numbers needs another route.

Can I compute share of voice and sentiment from DataForSEO's answers myself?

Share of voice is arithmetic you can write: your brand's mentions divided by every tracked brand's mentions over the same set of answers, after collapsing aliases, legal suffixes and product names onto one entity. Sentiment needs a classifier, which is a second billed call per answer. DataForSEO's own LLM Responses endpoints can run it, at $0.0002 per task on the standard queue plus the token cost charged by the model provider, where the worked example in its documentation for ChatGPT gpt-4.1-mini is $0.029631 for one task.

Which DataForSEO alternative covers Google AI Mode with both metrics?

DataForSEO's AI Optimization API has no AI Mode endpoint, so that surface has to come from elsewhere. API for GEO, our pick after hands-on testing by Mathias C., lists Google AI Mode among the six engines published on apiforgeo.com and returns share of voice, position, sentiment, ranked sources and per-brand perception for one credit, which is $13 per 1,000 answers down to $7 at its largest tier, pay as you go with no published minimum. LLM Pulse, AthenaHQ, Scrunch, OtterlyAI, GEOly, Ahrefs Brand Radar, SE Visible and Superlines also cover AI Mode with both metrics, at €30.67, $81.94, $500 a month, $15.75, $13.32, $20, $3.30 and €14.04 respectively.

Which of these alternatives let an agency resell the data?

Three put it in writing: API for GEO presents reselling as an intended use, LLM Pulse allows presenting its data to clients under your own branding on eligible plans while forbidding resale as a standalone dataset, and AthenaHQ's agency page offers discounted licences bundled with your managed services. Ahrefs, SE Visible, SE Ranking's Data API and Superlines forbid resale or redistribution in their terms. Scrunch, OtterlyAI and GEOly publish no clause we could read, which is not the same as permission.

Does DataForSEO's LLM Mentions API replace a share of voice?

It answers a different question. Its live target metrics endpoint takes up to ten domains or keywords and returns, per location, language, platform and source domain, a `mentions` count described as "total LLM mentions count" and an `ai_search_volume`, read from a database of 420,701,746 stored prompts. Those are counts over somebody else's prompt set rather than a share computed over your own tracked prompts, and the ChatGPT rows are limited to the United States in English. It costs $0.1 per task plus $0.001 per row.

Contents