DataForSEO has no subscription. You top up a balance (minimum $50), the balance never expires, and every API call deducts a fraction of a cent. New accounts get $1 of trial credit. That is the whole model. What the pricing page does not tell you is what real work adds up to, so here is my own account: 2,245 API requests and $34.42 for the 30 days ending 26 October 2026, covering every client site I touch, my own sites, and the AI-visibility studies I publish. Strip out the studies and the ordinary SEO work came to $17.37.
Disclosure: links to DataForSEO in this post are affiliate links. If you sign up through them, I may earn a commission on every successful sale, at no extra cost to you. I paid for DataForSEO for a year before joining their program, and every number below is from my own usage export.
The pricing model in sixty seconds
- Pay as you go. Top up at least $50. No monthly fee, no plan tiers, and the balance does not expire.
- Task plus items. Most endpoints charge a small fee per request (“task”) plus a smaller fee per row returned. Ask for 10 keywords and you pay for 10 rows. Ask for 1,000 and you pay for 1,000.
- Standard, priority, live. SERP requests come in three speeds. Standard is queued and cheapest. Live returns instantly and costs about three times as much. The MCP server uses live mode, because you are waiting in a chat window.
- Two multipliers to know. Asking a SERP call to load the AI Overview doubles that call. Asking a Labs call for clickstream data doubles that call.
- $1 trial. Enough to run the five test prompts from my MCP setup guide several times over.
The rates that matter for SEO work
DataForSEO has dozens of endpoints. These are the ones an SEO practitioner (not a software company) actually calls. Rates are from DataForSEO’s pricing pages on 27 October 2026.
| What you are doing | Endpoint family | Rate | Typical call |
|---|---|---|---|
| Keyword volume, difficulty, suggestions, related, ideas, competitors, ranked keywords | DataForSEO Labs (Google) | $0.012 per task + $0.00012 per item | 200 keyword suggestions: $0.036 |
| Search intent per keyword | DataForSEO Labs | $0.012 per task + $0.00012 per keyword | 100 keywords: $0.024 |
| Live Google results, top 10 | SERP API, live | $0.002 per SERP (+$0.002 per extra 10 results; +$0.002 to load the AI Overview) | One query, top 10, with AI Overview: $0.004 |
| Scheduled rank checks | SERP API, standard queue | $0.0006 per SERP | 500 keywords, once: $0.30 |
| Backlink summary, referring domains, anchors, new and lost | Backlinks API | $0.024 per task + $0.000036 per row | 1,000 backlinks: $0.06 |
| Site crawl | OnPage API | $0.00015 per crawled page | 2,000-page site: $0.30 |
| Page speed | OnPage Lighthouse | $0.005 per scanned page | 20 key pages: $0.10 |
| Google Ads search volume | Keywords Data | from $0.06 per task | One batch: $0.06 |
| Brand mentions in AI answers | LLM Mentions (AI Optimization) | $0.10 per task + $0.001 per row | 100 mentions: $0.20 |
| Raw AI answers (ChatGPT, Claude, Gemini, Perplexity) | LLM Responses (AI Optimization) | $0.0002 per task + the model’s own cost | Varies by model and answer length |
| Keyword volume inside AI tools | AI Keyword Search Volume | $0.01 per task + $0.0001 per row | 100 keywords: $0.02 |
Two rates on that table deserve a second look. The Backlinks API is where the gap to the subscription tools is widest: a thousand backlinks for six cents. And the Labs rate is the one you will pay most often, because keyword research, competitor research and ranked-keyword pulls all sit under it.
What 30 days of real work cost: my account, 27 September to 26 October 2026
This is the usage export from my dashboard, grouped by API. One account, every client, every research session, every study. I was active on 23 of the 30 days.
| API | What I used it for | Active days | 30-day cost | Share |
|---|---|---|---|---|
| AI Optimization | Pulling ChatGPT, Gemini, Perplexity and AI Overview answers for the AI-visibility studies I publish | 17 | $17.05 | 49.5% |
| DataForSEO Labs | Keyword research, competitor keyword gaps, ranked-keyword pulls for clients, my own sites, and this hub | 20 | $9.44 | 27.4% |
| Keywords Data | Google Ads search volume batches and trend checks | 11 | $3.97 | 11.5% |
| SERP | Live Google results for briefs, SERP checks, and the studies | 20 | $1.74 | 5.1% |
| Backlinks | Backlink profiles and referring-domain pulls for audits | 12 | $1.73 | 5.0% |
| OnPage | Crawls and page checks | 8 | $0.28 | 0.8% |
| Domain Analytics | Technology and WHOIS lookups | 2 | $0.19 | 0.6% |
| Merchant | One Amazon product lookup | 1 | $0.01 | 0.0% |
| Total | 23 | $34.42 | 100% |
Three things stand out in that table.
The studies cost more than the SEO. Half the bill is the AI Optimization API, which is what I use to collect hundreds of AI answers for research like the 105-answer study on who AI names as the best SEO expert in the Philippines. The three most expensive days in the whole month ($6.45 on 21 October, $5.17 on 8 October, $4.44 on 26 October) were study days or heavy keyword-research days, not client days. If you do not publish research, cut this line out and the month is $17.37.
A normal working day is under a dollar. The median active day was $0.86. Most days were a keyword pull or two, a handful of live SERPs, and the occasional backlink check. Eleven of the 23 active days cost under $0.60.
The SERP API is almost free at this volume. 702 live Google results, including the ones Claude pulls through the MCP server during briefs, came to $1.74. That is the line people expect to be expensive. It is the second cheapest.
The endpoints that actually cost money
The same export broken down by endpoint. 2,245 requests across 47 endpoints. These ten account for $29.57 of the $34.42.
| Endpoint | Requests | Cost | Per request | Why |
|---|---|---|---|---|
| llm_mentions/search | 27 | $6.10 | $0.226 | Each call returns up to hundreds of AI-answer rows at $0.001 each, on top of the $0.10 task fee. The single most expensive endpoint I use, and it only runs for studies. |
| ai_optimization/llm_responses | 418 | $5.78 | $0.014 | Full answers from ChatGPT, Gemini, Perplexity and Claude. The model’s own cost is passed through. |
| dataforseo_labs/ranked_keywords | 75 | $4.99 | $0.067 | Every keyword a domain ranks for. I pull 300 to 1,000 rows at a time, so the per-item fee adds up. |
| keywords_data/google_ads/search_volume | 41 | $3.69 | $0.090 | Google Ads volume batches. Flat $0.09 per task regardless of keyword count, so batch big. |
| serp/live/advanced | 702 | $1.74 | $0.0025 | Live Google results. Cheapest line per request on the list. |
| llm_mentions/top_domains | 17 | $1.72 | $0.101 | Which domains AI answers cite most for a topic. Study work. |
| ai_optimization/llm_scraper | 379 | $1.52 | $0.004 | Scraped chatbot answers with their citations, $0.0012 per page. |
| dataforseo_labs/keyword_suggestions | 53 | $1.50 | $0.028 | The workhorse of keyword research. About three cents a call. |
| llm_mentions/cross_aggregated_metrics | 12 | $1.21 | $0.101 | Share-of-voice style aggregates across AI engines. |
| dataforseo_labs/keyword_overview | 46 | $0.70 | $0.015 | Volume, difficulty and intent for a list of keywords. Under two cents a call. |
Everything else, 37 endpoints between them, added up to $4.85. That includes 145 instant page checks for $0.08, 135 content-parsing calls for $0.17, and 20 backlink summaries for $0.48.
A worked example you can check: the research behind this hub
On 26 October 2026 I ran the keyword research for this entire DataForSEO hub through Claude and the MCP server: 19 calls across keyword suggestions, related keywords, keyword ideas, a keyword overview for 85 terms, ranked keywords for four domains, and eight live SERPs. Priced at the list rates above, that session comes to about $0.31. The export for that day shows $0.43 on Labs and $0.03 on SERP, and that $0.46 includes a few unrelated pulls I made the same day. So the list-price estimate and the receipt agree within a few cents, and you can reproduce the session on the $1 trial credit.
Where it stops being cheap
Pay-per-call is a bargain for research and a trap for anything you run on a schedule without thinking. Rank tracking is the usual example, so here is the math for 500 keywords checked daily:
| Setup | Per check | Per month (30 days) |
|---|---|---|
| Top 10, standard queue | $0.0006 | $9.00 |
| Top 100, standard queue | $0.006 | $90.00 |
| Top 10, live mode | $0.002 | $30.00 |
| Top 100, live mode | $0.02 | $300.00 |
Top 10 on the standard queue is cheaper than any rank tracker I know of. Top 100 in live mode is more expensive than most of them. Same API, same keywords, a 33x difference, and the only variables are depth and speed. When you automate anything with DataForSEO, decide those two settings on purpose. The rank tracking post in this hub goes through how I set them.
The other two balance drains are in my own table. Clickstream data doubles every Labs call that supports it, so leave it off unless you need clickstream-adjusted volumes. And the LLM Mentions and LLM Responses endpoints bill per row and per model answer, which is why a study that pulls hundreds of full ChatGPT answers costs more than a month of keyword research.
Estimate your own month
Three numbers get you close enough to budget:
- Research calls per month (keyword, competitor, backlink pulls). Multiply by about $0.03. A typical call returns 100 to 300 rows.
- Live SERPs per month (briefs, spot checks). Multiply by $0.002, or $0.004 if you want the AI Overview each time.
- Tracked keywords × checks per month. Multiply by $0.0006 for top-10 standard queue.
Add them up. For one client with 60 research calls, 100 live SERPs and 200 keywords checked weekly, that is $1.80 + $0.20 + $0.48, about $2.50 a month. My own 30 days of non-study work across several sites came to $17.37, which lands in the same range per site. A $50 top-up covers one client for well over a year.
Questions people ask about DataForSEO pricing
Is DataForSEO free?
No. There is a $1 trial credit on new accounts and a free sandbox that returns sample data. Real data costs money per call. The minimum top-up is $50.
Does the balance expire?
No. DataForSEO’s pricing page states the balance never expires. Top up once and use it over months or years.
Is there a monthly plan?
No plans and no tiers. Every account pays the same per-call rates. High-volume users can ask for a custom quote.
What is the difference between standard, priority and live?
Speed. Standard is queued and cheapest ($0.0006 per Google SERP). Priority is queued but faster ($0.0012). Live returns in one request ($0.002). The MCP server uses live.
How much does DataForSEO cost per month for an SEO freelancer or small agency?
From my own export: $17.37 for 30 days of client and own-site SEO work, or $34.42 including published research, across 2,245 requests. A single client runs a few dollars a month. The number that moves the bill is how many full AI answers or deep SERPs you pull on a schedule, not how many keywords you research.
Can I see what each call cost?
Yes. The dashboard’s usage statistics break spend down by API, by endpoint and by day, and export to a spreadsheet. That export is where every number in this post came from.
Try it on the trial credit
Open a DataForSEO account (affiliate link), connect it to Claude with the setup guide, and run the five test prompts on the $1 trial. Then check the usage statistics and compare to my $0.31 session above. If the numbers make sense for your client load, the free skill pack turns those manual prompts into a repeatable month.
