Pangram Labs
AI text detector with segment-level evidence, plagiarism checking, and third-party verified accuracy.
For institutions that may have to justify an accusation, Pangram is the detector we'd put our name behind. AI Segment Analysis gives you the portion of a document that's machine-written, and Pangram publishes a 1-in-10,000 false positive rate rather than a single accuracy headline — both matter more than a confidence score when a student or contractor disputes the result. The Individual plan runs $20/mo billed annually (a $60/yr saving) or $20/mo month-to-month, and it includes plagiarism checking on every scan. Size the plan to your actual word count: 300,000 words a month goes fast in a large course, and Professional at $65/mo billed annually adds 5x the words plus $200/mo in API credits.
Verified 23h ago · liveness 80/100 · cite: rightaichoice.com/tools/pangram-labs
- University honor councils and professors who need segment-level evidence for an AI-detection ruling
- Publishing and media desks fact-checking contributor submissions and press releases at volume
- HR and recruiting teams screening application materials for machine-written text
- Developers wiring AI detection and plagiarism checks into submission portals via the Pangram API
- Anyone expecting a one-click verdict without reviewing segment highlights and evidence
- Production image moderation — the AI image detector is still a research preview with a 3-scan daily cap on free
- High-volume programs scanning millions of words a month on the $20/mo Individual tier without budgeting up to
We scan live Reddit threads, YouTube comments, X posts, G2 reviews and other communities — and hand you an honest verdict in under a minute.
- Honest verdict, not marketing
- Real pros & cons from real users
- Attributed quotes with receipts
3 free scans · no card needed
Skip Pangram if you need production-grade image moderation today — the image detector is a research preview capped at 100 scans a month on Individual, not a moderation pipeline.
API usage bills on top of your plan at $0.05 per 100 words for Pangram 4, so a high-volume integration costs far more than the seat price suggests.
Pangram sits above free web checkers and below institutional plagiarism platforms. Free covers 2,000 words/day for evaluation. Individual at $20/mo billed annually ($60/yr saving) or $20/mo month-to-month suits a single professor, editor, or recruiter at 300,000 words/month. Professional at $65/mo billed annually adds 5x words plus $200/mo API credits — the tier for a department or submissions desk. Enterprise and team pricing go through Contact Sales.
In short
Pangram Labs — AI text detector with segment-level evidence, plagiarism checking, and third-party verified accuracy. Best for University honor councils and professors who need segment-level evidence for an AI-detection ruling, Publishing and media desks fact-checking contributor submissions and press releases at volume, HR and recruiting teams screening application materials for machine-written text. Free to start; paid plans from $20/mo.
What's new in Pangram Labs
Checked todayAcross the latest 5 updates: 1 feature update, 1 launch, 1 changelog entry and 2 news mentions.
All About False Positives in AI Detectors
Pangram details how it reaches a false positive rate of 1 in 10,000 across its detection stack, the figure institutions cite when defending a detection call.
AI Helps Write Nearly 50% of Press Releases
Pangram scanned 178,672 PR Newswire releases and found AI-generated text in 34% over the past year, rising to 48% by June 2026.
Pangram 4 Migration Guide
A published guide for moving API integrations and workflows from Pangram 3 to Pangram 4, the company's current flagship detection model.
Introducing Pangram Image Detection (Research Preview)
Pangram launched a research preview of its AI image detection model for JPG, PNG, and WebP files, available alongside text detection.
Introducing Pangram 4
Pangram 4 released as the latest generation of the company's AI text detection model, with a companion technical overview comparing it to Pangram 3.
What people actually say about Pangram Labs — is it worth it?
We ran a structured research pass across product reviews, community discussions, and post-purchase forum threads to surface the patterns vendors won't publish themselves. Below: the recurring strengths, the hidden costs people mention most, and the cohort that consistently regrets adopting this tool.
26 mentions across 3 sources (Hacker News, YouTube, Lemmy) · researched Aug 5, 2026.
Average across the 3 sources that answered — each source counts once, not each post.
- +Research-driven approach with published papers (ICLR 2026) and academic collaboration
- +Low false-positive claims backed by independent university verification
- +Chrome extension detects AI in social media feeds in one click
- +Free tier supports up to 2,000-word file uploads
- +Multilingual detection in 20+ languages for global use
- −Flags 100% human-written text as AI-generated in community tests
- −Misses pure AI-generated text, labeling it human
- −99.98% accuracy claim is not replicated in real-world use
- −Free plan has word limits that may frustrate power users
- −No plugin available for WordPress or other CMS platforms
- • Free tier limited to 2,000 words may push users to paid plans quickly
- • Plagiarism checker only available on paid plans
- • Pricing details for API not publicly listed
Viability Score
How well maintained and how widely used is Pangram Labs? Built from what the vendor actually publishes (docs, changelog, tutorials, integrations, pricing), whether the site is live, and how much real users discuss it. How we calculate this
Last calculated: September 2026
How we score →Key Features
- Pangram 4 flagship AI text detection model, released July 2026
- Benchmarked against 26 models including Claude Opus 5 and GPT-5.6
- AI Segment Analysis showing how much of a document is AI-written
- AI Assistance Detection flagging segments edited by AI rather than generated
- Plagiarism checker bundled with every paid scan
- Reported false positive rate of 1 in 10,000 across the detection stack
- Multilingual detection across 20+ languages
- AI Image Detection research preview for JPG, PNG, WebP up to 30 MB
- Detects humanized or paraphrased text used to evade other detectors
- File upload and OCR for PDF, DOCX, RTF, up to 100 files at a time
- Chrome browser extension for checking AI text on any website
- Gmail integration for checking text in your inbox
- Google Docs add-on and LMS plugins for Canvas, Moodle, Google Classroom, Brightspace
- API access with realtime and bulk endpoints
- $200/mo API credits included on Professional
About Pangram Labs
Pangram Labs builds an AI content detector for people who have to defend the call they make. Universities, editors, recruiters, and compliance teams run documents through it to separate human writing from machine writing, and the company backs its numbers with third-party verification from researchers at the University of Chicago and the University of Maryland. The flagship model is Pangram 4, released July 29, 2026, benchmarked against 26 models from OpenAI, Anthropic, Google, Meta, DeepSeek, xAI, Mistral, and others; Pangram reports 99.7% accuracy across those AI-generated samples. What separates Pangram from a single-number checker is granularity. AI Segment Analysis shows how much of a document is machine-written rather than just a confidence score, and AI Assistance Detection highlights passages a human edited with AI rather than generated outright. A plagiarism checker runs on every paid scan, so copy-paste reuse and AI text surface in one pass. Pangram publishes a false positive rate of 1 in 10,000 across its detection stack and works across 20+ languages. Deployment happens where you already work: a Chrome browser extension, a Gmail integration, a Google Docs add-on, and LMS plugins for Canvas, Moodle, Google Classroom, and Brightspace. File uploads cover PDF, DOCX, and RTF with OCR for scanned documents, up to 100 files at a time. An AI Image Detector is live as a research preview for JPG, PNG, and WebP up to 30 MB. The free tier allows 2,000 words per day with 3 image scans, so you can pressure-test it on your own documents before paying.
Behind the Verdict
Pangram's argument is defensibility, and it's a real one. Where Turnitin and free web checkers hand you a percentage, Pangram hands you highlighted segments and a stated false positive rate — the difference between "this scored 87% AI" and "these three paragraphs were AI-generated, the rest your student wrote." That matters in any setting where the accusation has consequences: honor councils, editorial desks, legal review. The technical case is that Pangram trains on diverse datasets with hard negative mining and active learning rather than perplexity and burstiness scoring. You don't have to accept that framing on faith — the company publishes a model card and benchmark history, showing per-model accuracy for Claude Opus 5 (99.8%, tested July 2026), GPT-5.6 (99.6%, July 2026), Gemini 3 (99.5%, November 2025), Grok 4.3 (99.4%, July 2026), and a long tail from Claude Sonnet 5 to GLM 5.2. It reports 99.7% accuracy across AI-generated samples from all 26 benchmarked models. The company does original research, not just product marketing. Its own scans of 178,672 PR Newswire releases found AI-generated text in 34% over the past year, climbing to 48% by June 2026, and a separate scan of over a million social posts found AI content concentrated on LinkedIn. Those studies also function as evidence for the detector's usefulness. Where it doesn't fit: the AI Image Detector is still a research preview with a 3-scan daily cap on free, so nobody should be running production image moderation through it. And volume is the trap — 300,000 words a month sounds like a lot until a writing-intensive course or a submissions desk runs through it. Map your monthly word count before you pick a tier, and remember API usage bills separately at $0.05 per 100 words for Pangram 4. Against free web checkers, you pay for segment-level detail and a documented error rate. Against Turnitin and similar platforms, Pangram's pitch is that it does detection well rather than doing everything adequately, and it shows its work.
Researching Pangram Labs? Get your full AI stack in 60 seconds.
Free, no signup — tell us your goal and get tools matched to your budget & existing stack.
Real-world workflow fit
Concrete scenarios for the personas Pangram Labs actually fits — and what changes day-one when you adopt it.
A student's final paper reads oddly polished, so you upload the DOCX to Pangram and run the scan before the grade posts.
Outcome: AI Segment Analysis highlights the two AI-generated paragraphs while leaving the student's own analysis untouched, and the plagiarism checker clears the rest — you have segment-level evidence to bring to the honor council.
You install the Chrome extension and check incoming drafts as they land in Google Docs, then run a full plagiarism-plus-AI scan on anything that looks off.
Outcome: Every paid scan returns both the AI verdict and a plagiarism check in one pass, so you catch copy-paste reuse and machine-written copy without opening a second tool.
You wire the Pangram API into the upload flow so every student submission gets scored automatically, using the bulk endpoint for end-of-term volume.
Outcome: Scans run at $0.05 per 100 words for Pangram 4, and the $200 monthly API credits on the Professional tier absorb moderate departmental volume without a separate usage invoice.
Use Cases
- Teachers identifying AI-generated essays and paraphrased submissions
- Publishers verifying article originality at scale
- Content moderators detecting AI reviews and spam
- Developers scanning codebases and submission portals via API
- Law firms verifying legal documents
- Universities screening assignments and research
- Recruiters screening AI-generated resumes
- ML engineers filtering synthetic training data
Models Under the Hood
as of 2026-09-22
Limitations
- The free plan caps you at 2,000 words per day and 3 image detection scans daily.
- Individual is $20/mo billed annually (a $60/yr saving) or $20/mo month-to-month, covering 300,000 words per month and 100 image scans.
- Professional is $65/mo billed annually (a $240/yr saving) and adds 5x the words, 5x the image scans, and $200/mo of API credits.
- API usage bills separately at $0.05 per 100 words for Pangram 4 and $0.05 per 1,000 words for Pangram 3, with a 20% bulk discount.
- Image detection is a research preview supporting JPG, PNG, and WebP at minimum 512x512px up to 30 MB each.
as of 2026-09-29
Verification history
We have re-verified Pangram Labs 21 times since . Each pass re-reads the vendor's own pages and re-checks every listed field against that evidence; passes where nothing had changed are marked as such.
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
- — re-verified summary, description, our verdict, our analysis, pricing model, pricing tiers, features, integrations, who it suits, who should skip it
Showing the 6 most recent of 21 verification passes.
Free to cite with attribution — this page re-verifies continuously.
12-month cost
Project the real annual outlay, including the implied monthly cost when only an annual tier is published.
Vendor list price only. Add-on usage, seat overages, and contract minimums are surfaced under Hidden costs & gotchas.
Plans compared
For each published Pangram Labs tier: who it actually fits, and what it adds vs. the previous tier. Cross-reference the cost calculator above for projected annual outlay.
Free
$0/mo
Ideal for
A single professor, editor, or recruiter pressure-testing Pangram on their own documents before committing — no payment method required.
What this tier adds
Starting tier: 2,000 words per day, 3 image detection scans daily, file upload with OCR, 20+ language detection, Chrome extension, Google Docs integration, and API access.
Individual
$20/mo billed annually ($60/yr saving); $20/mo month-to-mont
Ideal for
One editor, professor, or recruiter running steady volume — around 300,000 words a month, or roughly a few hundred essay-length documents.
What this tier adds
Adds plagiarism detection on every scan, 100 image scans monthly, automated scanning of social media and Substack feeds, Gmail integration, and a 150x jump in monthly word allowance; $20/mo billed annually saves $60/yr.
Professional
$65/mo billed annually ($240/yr saving)
Ideal for
A university department, publishing desk, or small engineering team that needs both high scan volume and API bandwidth in one plan.
What this tier adds
Adds 5x the words and 5x the image scans over Individual, plus $200 of API credits monthly; $65/mo billed annually saves $240/yr.
Where the pricing makes sense
The company stage and team size where Pangram Labs's pricing actually pencils out — and where peers do it cheaper.
Pangram sits above free web checkers and below institutional plagiarism platforms. Free covers 2,000 words/day for evaluation. Individual at $20/mo billed annually ($60/yr saving) or $20/mo month-to-month suits a single professor, editor, or recruiter at 300,000 words/month. Professional at $65/mo billed annually adds 5x words plus $200/mo API credits — the tier for a department or submissions desk. Enterprise and team pricing go through Contact Sales.
Setup time & first value
How long it actually takes to get something useful out of Pangram Labs — broken out by persona, not the marketing-page minute.
Free tier: under two minutes — paste text or drop a file into the checker and read the segment highlights. Chrome extension: about a minute to install, then it checks text on any site you visit. Editor or developer: a few hours to wire the API into a submission portal, plus time to handle the realtime and bulk endpoint responses. Gmail and Google Docs take an OAuth connection each; LMS plugins
Switching to or from Pangram Labs
How to bring data in from common predecessors and how to get it back out — written for the switcher, not the buyer.
- →From Turnitin: run the same submissions through Pangram's checker to compare verdicts before you switch your grading workflow, since Pangram returns segment highlights rather than a single similarity score.
- →From GPTZero or a free web checker: paste the documents that mattered most and compare Pangram's AI Assistance Detection output against the confidence scores you were relying on.
- →From Pangram 3 to Pangram 4: follow the published Pangram 4 Migration Guide to move API calls over; Pangram 3 API pricing stays at $0.05 per 1,000 words if you need to run both during the transition.
- →From manual review: start with the Chrome extension and Gmail integration so checks happen in the tools your reviewers already open, rather than adding a new dashboard.
- ↗To Turnitin: if you need a single institutional platform covering similarity, grading, and peer review rather than detection alone, you'll be giving up segment-level AI evidence and Pangram's published 1-in-10,000 false
- ↗To a free web checker: fine for occasional curiosity, but you lose AI Segment Analysis, AI Assistance Detection, and the plagiarism check that runs on every paid Pangram scan.
- ↗To a self-hosted or open-source detector: relevant if data residency rules prevent sending documents to a third party, but you'd be trading a documented benchmark history for a model with no comparable published
Integrations
Resources & Guides
- Resourcepangram.com
AI Detector — Verified AI Content Checker | Pangram
Explore our blog for the latest on software updates, useful guides and AI news.
- Resourcepangram.com
AI Detector — Verified AI Content Checker | Pangram
Flexible pricing options depending on your AI detection needs. Choose between our individual, professional and enterprise pricing plans.
- Resourcepangram.com
Meet Pangram 3.3!
Introducing our latest AI detector, Pangram 3.3, with improved recall on humanized texts and an even lower false positive rate on ESL writing.
- Resourcepangram.com
Does Pangram work on GPT-5.4?
OpenAI just released a new flagship model: GPT-5.4! We break down if Pangram can detect it.
Tutorials & Learning
YouTube returned 6 videos for “Pangram Labs”, and we withheld 5: 5 could not be judged, because “Pangram Labs” is a single word that other videos use for other things. Showing the 1 we can prove is about Pangram Labs.
Official links
Tools that pair well with Pangram Labs
Common stack mates teams adopt alongside Pangram Labs, with the specific reason each pairing earns its keep.
Originality.AI
AI detection, plagiarism, fact-checking and readability in one content integrity suite.
Undetectable AI Pro
Undetectable AI Pro is a free AI humanizer built to make AI-generated text undetectable by AI detectors.
Netus AI
Netus AI is a two-app platform that humanizes AI text, detects AI writing free, and runs ChatGPT, Claude, Gemini, and Grok in one multi-model chat.
Alternatives to Pangram Labs
View all
Originality.AI
AI detection, plagiarism, fact-checking and readability in one content integrity suite.
Undetectable AI Pro
Undetectable AI Pro is a free AI humanizer built to make AI-generated text undetectable by AI detectors.
Frequently Asked Questions
Used Pangram Labs? Help shape our editorial sentiment research.