Skip to content
SPECIFICTION
Get listed

SlopTotal

sloptotal.com

Free plan Open source WebWindowsmacOSLinuxAPISelf-hosted
Visit sloptotal.com
Spec score 8.4 #17 of 45 in AI Content Detection Software

What it is

SlopTotal analyzes text and public URLs for signs of AI-generated writing. Users can paste text or submit a URL, which the service extracts into article content for analysis. Its 23 detection engines run in parallel using neural classifiers, statistical tests and linguistic heuristics. The report shows each engine’s vote and combines weighted votes into a final score, with results appearing as engines finish. A separate Site check reports whether a site appears to have been built with an AI app builder. The Free plan is free, with no account, credit card or scan limits; it includes API access and batch scanning but no report exports. Submitted text is processed on SlopTotal’s own servers, not sent to third-party detection APIs or LLM providers. Reports are temporarily retained and automatically deleted after 30 days. The detection backend is open source and can be self-hosted, and a Chrome extension can mark AI-looking results in Google Search and LinkedIn. The maker says scores are unreliable below about 80 words and settle at roughly 200 words and above, and should not decide grades, hiring or publication without human review.

Spec sheet

What it does
SlopTotal analyzes pasted text or public URLs with 23 independent AI-detection engines and combines their outputs into a calibrated score.Read 15 Oct 2026 · sloptotal.com
Privacy
The service says submitted text is not forwarded to OpenAI or other external detection APIs.Read 15 Oct 2026 · sloptotal.com
Support
For privacy questions, SlopTotal directs users to open an issue on the project’s GitHub repository or contact the maintainer by email listed there.Read 7 Oct 2026 · sloptotal.com
API access
The self-hosted API requires neither an account nor an API key.Read 15 Oct 2026 · sloptotal.com
API endpoints
Documented endpoints include full analysis, quick scoring, paragraph scoring, site scanning, file extraction and report retrieval.Read 15 Oct 2026 · sloptotal.com
Benchmark AUC
The published held-out benchmark reports an AUC of 0.942.Read 15 Oct 2026 · sloptotal.com
Benchmark size
SlopBench contains 3,252 texts split evenly between pre-ChatGPT human writing and outputs from 14 AI models.Read 15 Oct 2026 · sloptotal.com
Browser extension
The SlopTotal Chrome extension can mark AI-looking results in Google Search and LinkedIn.Read 7 Oct 2026 · github.com
Chrome extension
The extension can mark AI-looking content in Google Search and LinkedIn and can point to a selected SlopTotal backend.Read 15 Oct 2026 · github.com
Code limitation
SlopTotal states that its natural-language engines are not intended to detect AI-written source code.Read 15 Oct 2026 · sloptotal.com
Detection methods
The engine set combines neural classifiers, GPT-2-based statistical tests and linguistic heuristics.Read 15 Oct 2026 · sloptotal.com
Edited-text limitation
Humanized AI text was detected at a 55.4% rate in the benchmark's humanizing-prompt slice.Read 15 Oct 2026 · sloptotal.com
False positives
At the Likely AI threshold, 1.7% of benchmark human texts were classified as AI.Read 15 Oct 2026 · sloptotal.com
GPU support
CUDA is optional for faster inference; a GPU is not required.Read 15 Oct 2026 · github.com
Hosted API
The hosted service is available at api.sloptotal.com and uses a shared queue with 30-day report retention.Read 15 Oct 2026 · sloptotal.com
Human review
The maker says a detector score is evidence rather than a verdict and should not by itself decide a grade, hire or publication.Read 7 Oct 2026 · sloptotal.com
Input formats
The repository documents text, URLs and extraction from PDF, DOCX, TXT and Markdown files.Read 15 Oct 2026 · github.com
Licence
The application code is released under the MIT licence, while model weights retain their own licences.Read 15 Oct 2026 · github.com
Limit: human misclassification example
Korben reports a 2020 human blog post scoring 56.9 and receiving a Likely AI-generated verdict.Read 15 Oct 2026 · korben.info
Limit: independent test false positives
Korben's 36-text test classified 4 of 12 human samples as AI, all in English.Read 15 Oct 2026 · korben.info
Limit: independent test misses
In the same test, 9 of 24 AI-written samples were judged clean, all in French.Read 15 Oct 2026 · korben.info
Limit: sample result
Korben reports a 455-word Claude article scoring 23.6 and being labelled clean.Read 15 Oct 2026 · korben.info
Maker profile
The linked GitHub profile identifies the maker as Pablo Carrasco, a software engineer and computer-science graduate located in Valencia, Spain.Read 15 Oct 2026 · github.com
Non-native English limitation
15.4% of TOEFL essays by non-native English writers crossed the Likely AI threshold in the published evaluation.Read 15 Oct 2026 · sloptotal.com
Open source and self-hosting
The detection backend is open source and can be self-hosted on the user’s own hardware.Read 7 Oct 2026 · sloptotal.com
Per-engine transparency
Reports display each engine's score alongside the combined result.Read 15 Oct 2026 · sloptotal.com
Praise: local processing
Korben praised the ability to install SlopTotal locally and run 23 detectors without sending text to an outside provider.Read 15 Oct 2026 · korben.info
Praise: open development
SlopScore describes the repository as MIT-licensed, self-hostable, CPU-capable and backed by public accuracy measurements.Read 15 Oct 2026 · slopscore.org
Project contributors
SlopScore reports ten contributors to the repository.Read 15 Oct 2026 · slopscore.org
Release fact
SlopScore lists version 1.2.0 as released on October 10, 2026.Read 15 Oct 2026 · slopscore.org
Report retention
Analysis reports are stored temporarily and automatically deleted after 30 days.Read 7 Oct 2026 · sloptotal.com
Report scoring
The final score combines weighted engine votes, and results stream in as each engine finishes.Read 7 Oct 2026 · sloptotal.com
Retention
Hosted reports are automatically deleted after 30 days, with the original text not retained beyond the report excerpt.Read 15 Oct 2026 · sloptotal.com
Security posture
The privacy policy says the site uses no advertising cookies, fingerprinting or Google Analytics.Read 15 Oct 2026 · sloptotal.com
Self-hosting
A Docker image is published at ghcr.io/pablocaeg/sloptotal for running the service locally.Read 15 Oct 2026 · sloptotal.com
Short-text limitation
The documentation says results are weak below about 80 words and become more stable around 200 words.Read 15 Oct 2026 · sloptotal.com
Site check
The Site check feature reports whether a site appears to have been built with an AI app builder.Read 7 Oct 2026 · sloptotal.com
System requirements
The project documents Python 3.10 or newer, CPU operation, 4 GB RAM for its lite profile and roughly 2 GB for first-run model downloads.Read 15 Oct 2026 · github.com
Terms limitation
The terms describe results as probabilistic and prohibit using them as the sole basis for academic, employment, legal or disciplinary decisions.Read 15 Oct 2026 · sloptotal.com
Text inputs
Users can paste text or submit a URL, and SlopTotal extracts the article content automatically.Read 7 Oct 2026 · sloptotal.com
Tracking
The privacy policy says SlopTotal does not use cookies, analytics trackers or fingerprinting.Read 7 Oct 2026 · sloptotal.com
URL analysis
Public URLs are scraped for article text before detection, and the site-check mode can identify AI website-builder fingerprints.Read 15 Oct 2026 · sloptotal.com
Use cases
The maker describes use cases for educators, editors, SEO teams, recruiters, school teachers, law firms and e-commerce operations.Read 7 Oct 2026 · sloptotal.com

Plans and prices

Free Free No account · no credit card · no scan limits Read 15 Oct 2026 · sloptotal.com

Every SlopTotal plan and price →

Strong points and limits

Strong points

  • Uses 23 detection engines in parallel.
  • Shows each engine’s vote in the report.
  • Free plan has no scan limits or account requirement.
  • Backend is open source and self-hostable.
  • Automatically deletes reports after 30 days.

Limits

  • No report exports are offered.
  • Scores are unreliable below about 80 words.
  • Scores should not replace human judgment.

Who it suits

SlopTotal may suit educators, editors, SEO teams, recruiters, teachers, law firms and e-commerce operations that want an AI-writing signal. Its self-hostable backend may also suit users who want to run the detection system on their own hardware.

Verdict

SlopTotal makes engine-level votes visible alongside a combined score and offers a free, self-hostable option. Treat its results as evidence rather than a standalone decision, especially for short text.

Where Specifiction places it

Alternatives to SlopTotalAll alternatives →