What is OpenSEO? The open source SEO tool built on DataForSEO, explained

By the AgentMsgr team · Updated November 11, 2026 · 10 min read

OpenSEO is an open source SEO tool that you can use on its hosted website or run yourself for free. It covers keyword research, rank tracking, site audits, backlinks and tracking how AI chatbots mention your brand, and it charges for the data you actually look up instead of a big monthly subscription. Its data comes from DataForSEO. What sets it apart is that AI assistants like Claude and ChatGPT, and agents like Grok Bot, can connect to it and do SEO research for you.

What is OpenSEO, and who makes it?

OpenSEO is a product of Every App, Inc., and its website says the company is backed by Y Combinator. It pitches itself as the open source alternative to Ahrefs and Semrush. Its founder came to SEO from software engineering, wanted SEO data inside the AI tools they already used, and built the first version on top of DataForSEO.

It's new and it's growing fast. The code went up on GitHub in late February 2026. By November 11, 2026 the repository had about 27,000 stars and 3,800 forks, and version 0.1.27 came out that same day. US searches for "openseo" went from 10 a month at the start of 2026 to 4,400 in September and 5,400 in October, by Google Keyword Planner's count.

Is OpenSEO really an open source SEO tool?

Yes. The code is under the MIT license, one of the most permissive there is: you can read it, change it, and run your own copy. OpenSEO offers two ways to self-host: Docker on your own server or computer, or one deploy command to your own Cloudflare account, which works on Cloudflare's free plan.

Open source doesn't make the data free, though. A self-hosted copy needs your own DataForSEO API key, and you pay DataForSEO for every lookup.

Which OpenSEO is this? The older project with the same name

Search for OpenSEO and you may find a different, older tool. An open source project called OpenSEO from 2012 was built on the WebSharper framework, and its GitHub page says it's no longer maintained. Plenty of newer, unrelated repositories use the name too. The one this guide covers lives at github.com/every-app/open-seo and openseo.so.

OpenSEO features: keyword research, rank tracking, audits, backlinks and AI visibility

Here's what OpenSEO's docs list as of November 2026:

• Keyword research: ideas, search volume, difficulty, intent and cost per click, with the live Google results behind each phrase.
• Domain and competitor research: a domain's estimated traffic, the keywords it ranks for and its top pages.
• Backlinks: the links pointing at a site and the websites they come from.
• Rank tracking: positions on desktop, mobile or both, nationally or for a city, up to 1,000 keywords per tracker, checked daily, weekly or monthly.
• Site audits: a crawl that sorts problems like broken links and missing titles by severity. Up to 50 pages on the free plan and 10,000 on paid plans.
• AI visibility: saved questions checked in ChatGPT and Gemini by default, with Google AI Overviews, Claude and Perplexity as options, showing whether answers mention or link to you.
• Google Search Console and Analytics: your own clicks, impressions and organic traffic, read-only and free of credits.
• Local SEO: Google Business Profile checks, map results and map grid rank checks, through the agent tools.
• Project context and reports: your business, competitors and writing rules saved once, so every session starts from them, and research saved as reports.

OpenSEO pricing: plans, credits and self-hosting

OpenSEO pricing works like a phone plan with a data allowance. Each hosted plan includes a set amount of credits, shown in dollars, and every new lookup spends them. As of November 11, 2026:

On the hosted plans, each lookup costs what the data costs OpenSEO plus 28%. Its pricing page gives rough figures: about $0.05 per keyword search, $0.08 per backlink overview, up to 1.5 cents per keyword per rank check and about $0.002 per AI prompt check on one platform. Its own example month, 100 keyword searches, 20 backlink checks and 50 keywords tracked weekly, comes to $9.98, inside the Base plan.

What would OpenSEO cost you each month?

Put in a normal month of your own work. The defaults are that same example month, and the math uses the rounded per-lookup prices above, so it lands a few cents under OpenSEO's $9.98. Treat it as a ballpark: repeat lookups in a project are free for a while, which can pull the real bill lower.

A few more details from the billing docs. A yearly plan costs the same as 10 months. Extra credits come in $10 to $99 top-ups that never expire, while the monthly allowance resets each cycle. Saved results, exports, reports and your Search Console data don't use credits. Subscriptions have a 30-day money-back guarantee, which doesn't cover bought credits.

What powers OpenSEO? DataForSEO, explained

DataForSEO is a company that sells SEO data through an API, a way for one piece of software to request data from another. It has no subscriptions: you add money and pay per request, with a $50 minimum top-up, a $1 trial credit for new accounts and balances that don't expire. A Google results page, for example, starts at $0.0006. It also has its own MCP server.

OpenSEO says its keyword, results page, ranking, backlink, local and AI visibility data all come from DataForSEO, while your Search Console and Analytics data comes from Google. It's open about the deal: "It's the same data," its comparison page says. What it adds on top:

• Smaller answers for agents. Keyword rows drop the 12-month trend unless the agent asks, so the agent's working memory doesn't fill with numbers it won't use.
• Prices up front. Each agent tool tells the agent roughly what a call costs before it makes it.
• Free repeats. The same lookup in the same project is reused for a while: 24 hours for keyword research, 12 for results pages, 6 for backlinks.
• Cheaper scheduled rank checks. OpenSEO says it sends them through DataForSEO's task queue, which costs about a third of the live price.
• Everything around the data: rank and prompt tracking, saved project context, reports and ready-made skills.

You can keep DataForSEO's own MCP connected alongside OpenSEO for anything OpenSEO doesn't cover.

OpenSEO MCP: connecting Claude, ChatGPT and other agents

MCP, short for Model Context Protocol, is an open standard for plugging an AI assistant into another app so it can read the app's data and take actions there. The OpenSEO MCP server, at https://app.openseo.so/mcp, lets your assistant run keyword research, check results pages, read your Search Console data, set up rank trackers and save reports, all from a chat.

You connect in one of two ways. The usual one is signing in: the first time, you log in to OpenSEO and approve access. The other is an API key, which starts with "oseo_", is made under Settings, then API keys, and is shown only once. OpenSEO's docs cover:

• Claude on the web, desktop, phone, Claude Code and Cowork. The connector in Claude's directory works on Claude's free plan; the plugin with OpenSEO's skills needs a paid plan.
• ChatGPT on free and paid plans, Codex and Cursor.
• Grok Bot and Hermes, through a setup prompt you paste into the agent.

Skills are ready-made workflows you start with a slash command, such as /seo-audit or /competitor-analysis, and each one ends with a saved report. One design choice is worth knowing: when an agent adds keywords to a rank tracker, OpenSEO doesn't start a paid check on its own. It asks you to approve the cost first. For the Search Console side by itself, see Google Search Console MCP, and for GA4 data, see Google Analytics MCP.

OpenSEO vs Ahrefs and Semrush: which to pick

OpenSEO vs Ahrefs

Ahrefs runs its own crawler, which it calls the second most active on the web, and builds its own index from it. Its main plans start with Lite at $129 a month, and it lists API and MCP access on Lite and up. OpenSEO has no index of its own and starts at $10 a month plus what you use. If you look things up a few times a week, OpenSEO costs far less. If you dig through backlinks all day, Ahrefs' own data and flat price may suit you better.

OpenSEO vs Semrush

Semrush's SEO plan is $139 a month billed monthly, or $117.33 a month billed yearly, and its Starter plan at $199 a month adds MCP access. Its free plan allows 10 reports a day. Semrush also bundles AI visibility tracking and a much wider set of marketing tools. OpenSEO is narrower and cheaper, and it was built around being used through an agent.

OpenSEO review: who it suits and where it falls short

This OpenSEO review is based on its website, docs, pricing page and code, read on November 11. It suits solo site owners, small agencies and anyone who already works with AI agents and wants real SEO data inside them. The limits worth knowing:

• It's young. Version 0.1.27, released less than nine months after the code went public. Features and prices can change.
• The data is only as good as DataForSEO's. Volumes and backlink counts are estimates, as they are in every SEO tool.
• Prompt Research covers US English only, and its questions come mostly from Google's People also ask boxes, not from real chatbot logs.
• The free plan is a trial. No rank checks, no Prompt Research, and the $0.50 doesn't refill.
• Costs can creep. Daily tracking costs 7 times weekly, and when credits run out, scheduled checks are skipped.
• Self-hosting is real work. You run the updates, set up your own Google sign-in to connect Search Console, and on Cloudflare's free plan audits stop at about 40 pages.

OpenSEO with more than one agent: research, draft, approve

An OpenSEO workflow splits well across agents in one AgentMsgr group chat. One agent with the OpenSEO connector, a Grok Bot for example, pulls 20 low-competition keywords and checks who ranks for each. A Hermes agent, the other agent OpenSEO's setup page walks through, turns the top three into a page outline. A third agent reviews the draft against your notes before you see it.

You pass the work along with @mentions and keep the money decisions for yourself. Ask the agents to post "Approval needed:" before they spend credits or publish, and that message arrives with Approve and Decline buttons. It lines up with OpenSEO's own habit of asking before a paid rank check. To put the data to work, start with how to find low competition keywords or a keyword gap analysis.

Sources: OpenSEO's website, pricing, about page, OpenSEO vs DataForSEO and docs on billing, MCP, MCP tools, agent setup, rank tracking, site audits, AI visibility and self-hosting; the every-app/open-seo repository on GitHub (README, license and releases); the older 2012 OpenSEO repository's GitHub page; DataForSEO pricing; Ahrefs pricing; Semrush pricing. Search volumes from Google Keyword Planner (US, English, through October 2026). All read November 11, 2026. No affiliate links here, and AgentMsgr has no tie to OpenSEO, Every App, DataForSEO, Ahrefs, Semrush, SpaceXSI (formerly xAI) or Nous Research.

OpenSEO, the fine print