Skip to content
Metric VaultHelp Center
Open app

AI Article Generator

Four long-form content workflows behind one input — SEO brief, article outline, a full written draft, and a content-gap comparison against a competitor.

Last updated 2026-08-06

Summary#

AI Article Generator is one screen with four tabs, each running a different tool against the keyword in the input: an SEO brief, an article outline, a full written draft, and a content-gap comparison. Costs range from 1 to 3 credits depending on which tab is active.

Purpose#

Writing a page that ranks takes four separate pieces of work: understanding the keyword, deciding the structure, producing the words, and checking what a rival covers that you do not. Doing each in a different tool means retyping the keyword four times and losing the thread between them.

This screen puts all four behind one field. The decision it supports is what to publish next and what that page has to contain.

Overview#

The tab strip rewrites what the run button does. Switching tabs changes both the tool that runs and the label on the button — the input field stays as it is.

TabRunsButton becomesActual cost
SEO Brief Generatorseo_briefGenerate Brief · 1 report1 credit
Article Outlinecontent_templateGenerate Outline · 1 report1 credit
Full Article Writerarticle_writerWrite Article · 3 reports1 credit
Content Gap Analysiskeyword_gapAnalyze Gap · 1 report3 credits

Two of those button labels do not match the charge. Write Article · 3 reports costs 1 credit, and Analyze Gap · 1 report costs 3. Credit cost table is the authority; the labels are not.

The first three tabs are grounded on the same live data: related keyword suggestions, the real Google results for the term read to a depth of 30 with their People-Also-Ask questions, and a measured difficulty score. The fourth tab is a different kind of tool entirely — it compares two domains' ranked keywords rather than writing anything.

Benefits#

  • One input drives four workflows, so a keyword is typed once.
  • The draft arrives structured — TL;DR, H2 sections, key takeaways, FAQ, conclusion, call to action — rather than as a wall of text.
  • Copy buttons hand you HTML, Markdown or plain text for your CMS.
  • The draft carries its own title tag and meta description with character counts.
  • The brief and outline tabs cost 1 credit each, below the monthly premium threshold.

Use Cases#

Producing a first draft fast. The Full Article Writer returns a complete short-form article, typically 600 to 800 structured words, in about a minute.

Scoping before writing. Run the brief tab, read the difficulty and the competing pages, and decide whether the page is worth writing at all.

Structuring an existing idea. The outline tab gives the heading skeleton and the questions to answer.

Finding what a competitor covers. The gap tab lists the keywords they rank for and you do not.

Feeding the rest of the content chain. Take the draft into Content Optimizer to score it, or Content Repurposer to spread it.

Requirements#

  • A signed-in account.
  • A paid plan; no tab costs zero, and any cost above zero is blocked on Free.
  • Enough monthly credits for the Content Gap tab, which at 3 credits is metered against your monthly quota. The other three tabs are not.
  • For the Content Gap tab, two domains in the field separated by vs.

Permissions#

Any workspace member on a paid plan can run every tab. Free accounts are refused with This tool needs a paid plan. Free includes the 10 technical SEO tools; upgrade to Pro to unlock the rest. Signed-out visitors see Please sign in to run this tool. The recommendations panel under a result requires Pro or above.

Cost#

Per run, by tab: SEO Brief Generator 1 credit, Article Outline 1 credit, Full Article Writer 1 credit, Content Gap Analysis 3 credits.

The first three sit below the premium threshold of 3, so they are limited only by the hourly fair-use cap of 100 light-tool calls per hour. Content Gap Analysis meets the threshold and is metered against your monthly quota.

None of the first three is served from the shared cross-customer cache, so each run is fresh. Content Gap results are cached across customers for one day, and a cache hit still charges the 3 credits. If you have run the same query before, the You already ran this dialog offers Open saved result, which is free. See How credits work and Result caching and freshness.

Dashboard → Write & Optimize → AI Article Generator

Inputs#

FieldAcceptsRequiredDefaultValidationNotes
e.g. email marketingA keyword or topic. On the Content Gap tab, yourdomain.com vs competitor.comYesEmptyEmpty shows Enter a value first.The same field serves all four tabs. Enter runs the active tab
Tab stripSEO Brief Generator, Article Outline, Full Article Writer, Content Gap AnalysisYesSEO Brief GeneratorSwitching a tab rewrites the run button and the tool behind it
Try sampleNoRuns the active tab with whatever is in the field; it does not insert an example
Country picker (topbar)Country listNoUnited StatesRecorded with the run, but the grounding data always comes from the United States, English-language index
Date range (topbar)Hidden on this view

Step-by-Step Guide#

  1. Open Write & Optimize and choose AI Article Generator.
  2. Read the How It Works strip: brief a keyword, outline the H2 structure, write the full article.
  3. Under Choose a Workflow, click the tab you want. The run button relabels itself.
  4. Type your keyword into the field marked e.g. email marketing. On the Content Gap Analysis tab, type yourdomain.com vs competitor.com.
  5. Click the run button, or press Enter.
  6. Wait through Analyzing… this can take a few seconds for live data. A full article takes longest — allow around 90 seconds.
  7. On a draft, use Copy as HTML, Copy as Markdown or Copy plain text to move it into your CMS.

Reading the Results#

Full Article Writer#

The header. The article's own title is the headline, followed by a Draft ready badge, the target keyword, the word count and the reading time, with an Article pill on the right.

The four KPI tiles.

TileSub-labelReadsGoodBadAction
Word countfull draftWords in the draft600-800 for this toolUnder 300Well under 300 means the model returned a stub — re-run
Reading timeminutesEstimated at roughly 220 words a minuteMatches your audienceCompare with the reading time in the brief
SectionsH2 blocksTop-level sections written5-80-2Two sections is a fragment, not an article
FAQPAA answersQuestion-and-answer pairs appended3 or more0No FAQ means no answer-box coverage; add the brief's questions

Use this draftcopy into your CMS. Three buttons: Copy as HTML, Copy as Markdown and Copy plain text. The HTML version wraps the title in an h1, the sections in h2, bullets in lists, and marks the TL;DR and key takeaways in bold. The Markdown version uses heading levels and blockquotes for the takeaways. Use HTML for a WordPress-style editor and Markdown for a static site.

SEO metadatatitle tag and description. Three rows. Title tag shows its length out of 60. Description shows its length out of 160. Secondary kws lists the supporting keywords. Both counters color by fit, so treat anything over the limit as an edit before publishing.

The articleN words · N min read. The draft itself, in reading order: the H1, a TL;DR block, the introduction, then each section as an H2 with its paragraphs, an optional bullet list and a Key takeaway callout. After the sections come Frequently asked questions, a Conclusion and a call-to-action block. Read the key takeaways first — if they are all saying the same thing, the draft is repeating itself and needs cutting.

Best practices applied. Chips naming the writing conventions the model followed. Useful as a checklist when you edit.

If the writer returns nothing, the article block says: The writer returned no article body. Try running it again — if it keeps happening, shorten the topic.

SEO Brief Generator and Article Outline#

These two tabs render the same reports as the tool on the sidebar row above them. The brief gives Search Intent, Target Words, Difficulty and Competing Pages, then title options, a meta snippet, the outline, on-page targets, the People-Also-Ask questions and the competing pages. The outline tab returns the scored content template — target words, difficulty, top competitors, PAA count, recommended headings, semantic cluster, formats and a 30-day plan. Both are documented in full in Content Brief & Template.

Note that the Article Outline tab runs the content-template tool, so its output is a scored on-page specification rather than a bare list of headings.

Content Gap Analysis#

This tab compares two domains rather than writing anything. It needs both halves of yourdomain.com vs competitor.com; with only one domain it returns the honest no-data card instead of a report. The result covers the keywords the competitor ranks for that you do not, the ones where you are weaker, the ones where you are stronger, the size and estimated value of the gap, and a set of quick wins. Gap Finder documents every block of it.

What to do with it. Brief first, outline second, write third, then score the published page with Content Optimizer and check it with Pre-Publish Trust Checks before it goes live.

Examples#

Example

Example: You need a post on email deliverability. On the SEO Brief Generator tab you spend 1 credit and learn the target is 1,800 words at difficulty 52, with 11 People-Also-Ask questions. You switch to Full Article Writer, leave the keyword in place, and spend another credit. Ninety seconds later the draft header reads 740 words, 4 minutes, 6 sections and 5 FAQ answers. The SEO metadata block shows a 58/60 title and a 152/160 description, both in range. You copy the draft as Markdown, expand it toward the brief's 1,800-word target using the questions the brief listed, and publish.

Screenshots#

Screenshot
the AI Article Generator view with the four-tab strip under "Choose a Workflow" and the run button reading "Generate Brief · 1 report"
Screenshot
the same view with the Full Article Writer tab active and the button reading "Write Article · 3 reports"
Screenshot
an article result header showing the title, Draft ready badge and the Word count, Reading time, Sections and FAQ tiles
Screenshot
the "Use this draft" copy buttons above the SEO metadata rows with the title-tag and description character counts

Tips#

  • Check the button label before you click. The tab strip is the only thing that tells you which tool is armed.
  • Run the brief tab first and keep it open. The draft is easier to expand when you have the target length and the questions on screen.
  • The writer produces short-form deliberately. Treat the draft as a strong skeleton, not a finished 2,000-word piece.
  • On the Content Gap tab remember the vs — without it the run returns nothing useful and still costs 3 credits.
  • Use Copy as Markdown if your CMS accepts it; it survives paste far better than rich text.

Best Practices#

  • Always edit the draft. Add your own data, examples and internal links before publishing.
  • Keep the FAQ block — it is built from real People-Also-Ask questions and is the cheapest answer-box coverage you will get.
  • Score every published draft with Content Optimizer against the same keyword.
  • Run Pre-Publish Trust Checks on anything model-written before it goes live.
  • When you want the whole chain in one action, use Auto-Execute, which runs brief, outline, article and optimizer in sequence.

Common Mistakes#

  • Believing the button labels. Write Article · 3 reports costs 1; Analyze Gap · 1 report costs 3.
  • Expecting the Article Outline tab to run an outline-only tool. It runs the content template.
  • Publishing the draft unedited. It is short-form and un-fact-checked.
  • Typing a single domain on the Content Gap tab. Both sides are required.
  • Switching tabs mid-run. Let the current run finish; the button is disabled while it works.

Limitations#

  • Pick a length before you generate: 800, 1,500 or 2,500 words. The target is structural — a longer article is written as more H2 sections, not as the same outline padded — and the draft header reports the words it actually produced against the target you asked for, so a shortfall is visible rather than silent.
  • 2,500 is the ceiling, and the reason is the platform rather than the model: Cloudflare closes a connection that has been silent for over 100 seconds, so the generation budget is capped at 90.
  • Grounding data comes from Google's United States, English-language index; the country picker does not change it.
  • Statistics are sourced-claims-only. The model may state a number only if it cites one of the live top-ranking pages fetched for your keyword, and any claim citing anything else is removed before you see the draft — the header tells you how many were dropped. Advice and method carry no citation, because they need none.
  • What that does and does not prove: a cited claim points at a real page that really ranks for your keyword. It is not a fact-check. Metric Vault reads each result's title and description, not its body, so it cannot confirm the page contains the number. Treat a citation as a checkable lead, not a verification.
  • The article body is generated in a single pass, so it does not read your existing site or match your house style unless you edit it.
  • The Article Outline tab does not run the outline tool that exists behind the scenes; that one is only reached through Auto-Execute.
  • Very long or unusual topics are the ones most likely to return an empty body.

Troubleshooting#

SymptomLikely causeFix
Enter a value first.Empty fieldType a keyword
Please sign in to run this tool.Signed outSign in and run again
This tool needs a paid plan…Free planUpgrade — see Plans and pricing
The writer returned no article body. Try running it again — if it keeps happening, shorten the topic.The generation returned nothing usableRe-run, then shorten or simplify the topic
The draft is much shorter than expectedExpected — this tool writes short-formExpand it using the brief's questions
Content Gap returns an empty cardOnly one domain was enteredUse yourdomain.com vs competitor.com
The wrong report appearedA different tab was activeCheck the tab strip and the button label, then re-run
You already ran this dialog appearsA matching earlier run existsOpen saved result is free; Run fresh charges again
Monthly limit reached — upgrade to continue.Out of credits, usually on the 3-credit gap tabSee I ran out of credits
The run takes over a minuteNormal for the Full Article WriterWait — the screen estimates 90 seconds

FAQs#

How long is the article it writes? Short-form: roughly 600 to 800 structured words with a TL;DR, several H2 sections, key takeaways, an FAQ and a conclusion. It is deliberately capped, because longer generations were not completing reliably.

Why does the Full Article Writer button say three reports if it costs one? The label was not updated when the price was set. The charge is 1 credit. Credit cost table lists the real cost of every tool.

Does the article come with images or citations? No. It returns text, a suggested title tag, a meta description and secondary keywords. Add media, sources and internal links yourself.

Which tab should I use if I only have budget for one run? The brief tab. Knowing the target length, difficulty and questions makes a human-written page better than a generated one written blind.

How do I get a longer piece? Generate the draft, then expand it against the brief's target word count and its question list. Or run Auto-Execute, which chains brief, outline, article and optimizer for four credits total.

Where does the draft go afterwards? Every run is saved automatically and appears in Saved Work and My Content. Re-opening a saved result costs nothing.

See also

Was this article helpful?