Open-source, self-hosted SEO dashboard that runs on –$4/mo of API credits

A developer released SEO Command Center, an open-source, self-hosted SEO dashboard that runs on DataForSEO's pay-as-you-go API for roughly $4 per month in API credits, undercutting subscription-based tools like Ahrefs and Local Falcon. The platform offers rank tracking, keyword research, competitor gap analysis, AI Overview visibility, site audits, link gap analysis, and local map-pack grids, all without subscription fees or keyword caps.

A self-hosted SEO platform that costs ~$4/month instead of $139/month. Rank tracking · keyword research · competitor gaps · AI Overview visibility · site audits · link gaps · local map-pack grids SEO tools charge subscription prices for data they buy wholesale. Under the hood they're calling the same SERP APIs you can call yourself. SEO Command Center is the whole dashboard, self-hosted, on top of DataForSEO https://dataforseo.com/?aff=254685 's pay-as-you-go API. No subscription, no seat pricing, no keyword caps — you pay cents for the exact SERPs you pull: | SEO Command Center | | |---| Ahrefs Lite https://ahrefs.com/pricing Local Falcon https://www.localfalcon.com/pricing pay-as-you-go ~$4/mo typical Measured on a real deployment: 570 keywords across 5 sites updated daily, plus weekly local map grids and brand-SERP checks — $3.55/month in API costs. A 7×7 map-grid scan is $0.29; a keyword-research pull is ~$0.02. Seven pages, one dark command center, everything driven from the browser: | Page | What it does | |---|---| Rankings | Daily positions, 1d/7d/30d deltas, best-ever, sparklines, history charts, visibility score, GSC impressions/clicks, mobile ranks, CSV export | Research | Live keyword ideas from seeds or any competitor's domain — volume, CPC, competition — checkbox → tracked | Competitors | Who actually outranks you + every keyword they rank for that you don't | AI Visibility | Which of your keywords trigger Google AI Overviews, whether you're cited, and who is | Site Health | Zero-cost technical crawler: broken links, duplicate titles, missing metas, thin content, per-site score | Link Gap | Domains linking to your rivals but not you, junk-filtered and classified by how to actually get the link | Map Grid | LocalFalcon-style GPS grid: your Google map-pack rank from 49+ points around your city, plus who holds the top 3 at every pin | More screenshots in docs/screenshots /testedmedia/seo-command-center/blob/main/docs/screenshots . Try it with sample data first no account needed : git clone https://github.com/testedmedia/seo-command-center.git cd seo-command-center python3 demo.py && python3 worker.py serve → http://localhost:8000 Track your own site: python3 setup.py guided: DataForSEO creds validated live , your sites, options python3 worker.py run all first pull — typically $0.10-0.50 python3 worker.py serve local dashboard That's it. Python 3.9+ standard library only — no pip install, no Docker, no database server SQLite file . | Requirement | For | Cost | |---|---|---| | Python 3.9+ | everything | free | | worker.py deploy wrangler Local mode is read-only pages. Hosted mode free Cloudflare Pages unlocks the buttons: refresh any report, add keywords/sites, delete keywords, live research, and map-grid setup from the browser — protected by a login page orange matrix included . python3 worker.py deploy-config creates the Pages project + queue, pushes your secrets python3 worker.py deploy ship the dashboard python3 worker.py loop poll the queue + daily auto-refresh cron/systemd/launchd Buttons queue work into Cloudflare KV; worker.py loop on any machine a laptop, a $4 VPS, a Raspberry Pi picks jobs up within 2 minutes, runs them, and redeploys. Your API credentials only ever live in your .env and as encrypted Pages secrets. - Create an OAuth client Desktop in Google Cloud Console https://console.cloud.google.com/apis/credentials , enable the Search Console API. - Save the client JSON to credentials/gsc-client-secret.json . - Run any OAuth helper to get a refresh token into credentials/gsc-tokens.json {"refresh token": "..."} . - That's it — rankings runs will merge impressions, clicks and Google-reported positions, and auto-discover keywords you already get traffic for. ┌────────────────────┐ ┌──────────────────────┐ ┌─────────────────────┐ │ DataForSEO API │ ←── │ tracker/ .py │ ──→ │ data/ranks.db │ │ SERPs, keywords, │ │ stdlib Python, │ │ SQLite, all │ │ maps, backlinks │ │ one file per report │ │ history forever │ └────────────────────┘ └──────────┬───────────┘ └─────────────────────┘ │ renders ▼ ┌──────────────────────┐ ┌─────────────────────┐ │ site/ .html │ ──→ │ localhost or │ │ self-contained dark │ │ Cloudflare Pages │ │ dashboard pages │ │ + auth, buttons │ └──────────────────────┘ └─────────────────────┘ No framework. Each report is one Python file that pulls data, stores history in SQLite, and renders a self-contained HTML page. No backend server. Hosted interactivity is three tiny Cloudflare Pages Functions auth, queue, research proxy . No lock-in. Your entire SEO history is one SQLite file you can query directly. - Manual refreshes are capped per tool per day default 2 and research pulls per day default 20 — a shared dashboard can't drain your balance. - Every research pull and grid save shows its dollar cost before/after running. - The keyword cap per site default 100, configurable bounds the daily tracking spend. - Weekly-only for the expensive stuff map grids, true-SERP brand checks by default. Is the ranking data as good as Semrush/Ahrefs? It's the same class of data — DataForSEO is one of the SERP providers this industry runs on. Tracked positions come from live Google SERPs top 100, any country/language . Can I track competitors' sites too? Yes — add any domain as a site, or use Research → "Steal from a competitor" and the Competitors gap report. Multiple sites? Unlimited. Cost scales with keywords, not sites. White-label / client dashboards? Out of the box it carries the Tested Media logo. Set BRAND NAME and optionally LOGO FILE in .env to rebrand everything — sidebar, login page, titles. Deploy once per client, each with its own access key. Why not just use rank tracker X? If you're happy paying monthly SaaS pricing, do that. This is for people who want the whole stack, their own data, and a bill that rounds to a coffee. PRs welcome. The codebase is intentionally boring: stdlib Python, one file per report, no build step. Add a report = one script that writes JSON + HTML, one nav entry.