Free with an account · Meta · Takeaways · Pull quote · Conclusion
Title plus optional outline. Pick length and tone. Get a publishable draft — SEO meta, takeaways box, hook intro, H2 sections, pull quote, conclusion.
See how it works — click any example
Short 500-800 words / medium 900-1,400 / long 1,500-2,200. Tones: informative, conversational, professional, casual. Keywords (1-4) optional but sharpen the draft. ~30-130s per run.
Full draft with SEO meta title + description, key-takeaways box, hook intro, 3-8 H2 sections (each with a takeaway pill and pull quote), conclusion, and 4-8 internal-link-anchor suggestions. Start with a Hook or Ideas; polish with the Humanizer.
Claims that need hard data we don't have — bring the numbers. Niche-specific facts we can't verify — add context. Drafts over 3,500 words — split into a series. "Revolutionary", "leverage", "in today's fast-paced world" — 20+ AI-slop phrases stripped.
Free with an account — 30/month. Upgrade to Pro for 1,500/month. Each draft costs 4 credits (full-length generation).
Input: "How to choose a standing desk converter that actually fits your desk" · Medium · Conversational · 1,150 words
KEY TAKEAWAYS
Most standing desk converter returns come down to one mistake: the buyer never measured the desk. A 36-inch-deep desk will happily host a standard converter; a 24-inch-deep apartment desk usually cannot. Before you look at brands, prices, or electric-vs-spring, spend two minutes with a tape measure. The right converter for your desk is often the cheapest one that physically fits — and if your desk is too shallow, no amount of feature comparison will rescue the setup. Here is the decision order that matches how buyers who kept their converters worked through it.
TAKEAWAY · Depth matters more than width — measure your desk before shopping.
Measure your desk surface first — depth is the real constraint
Most converters are between 30 and 36 inches wide, but the number that sinks setups is depth. A converter needs roughly 20-24 inches of depth to sit flat without your keyboard dangling or your monitor too close to your face. Measure from the front edge of your desk to the wall. If you are under 24 inches, look at compact models or a wall-mounted arm instead. Also measure clearance behind the desk — a riser that extends backward when raised can hit the wall, and spring-assist models sometimes need 6-12 inches of swing room.
TAKEAWAY · Spring-assist caps around 35 lbs — dual or ultrawide monitors push you toward electric.
Match weight capacity to your monitor setup
Single 24-27" monitor on a standard arm: spring-assist is fine and cheaper. Dual 27" or a single 34" ultrawide: you are probably past the spring range. Most spring-assist converters comfortably handle 25-35 lbs of top-deck load; above that, the mechanism either fails to lift or bounces. Electric models (typically $250-450) handle 40-60 lbs without drama and add quieter adjustment — the real upgrade is not the weight capacity, it is the "just press the button" ergonomic habit. If you adjust less than once a day, skip the electric premium.
+ 3 more sections · pull quote · 4-8 internal-link anchors · conclusion · compliance flags
Paste a working title — that is the only required input. Optionally paste an outline, describe your audience, list your target keywords, pick a length (short 500-800 words / medium 900-1400 / long 1500-2200) and a tone (informative / conversational / professional / casual). In roughly 30-130 seconds you get a complete publishable draft: SEO meta title (≤60 chars) and meta description (110-160 chars), a refined final title, a hook introduction (80-150 words, one paragraph), 3-6 key-takeaways bullets for a highlighted box at the top, 3-8 H2 sections each with a 150-600 word body and a one-sentence takeaway pill, a pull quote, a forward-looking conclusion (80-150 words), 4-8 internal-link-anchor suggestions, word count, reading time, and any compliance flags. It is a draft — expect 10-20 minutes of editing before publishing. It is NOT a research tool — more on that below.
Two real differences. First, every draft comes back in a fixed JSON schema — we enforce what a good blog post contains (meta, key takeaways, section-level takeaways, pull quote, internal-link anchors, word count) instead of hoping you get them. You cannot forget to ask for a meta description. Second, we engineer the constraints: twenty-something banned phrases stripped out ("revolutionary", "cutting-edge", "unlock", "supercharge", "in today's fast-paced world", "leverage", "delve into"), no "Let's dive in." openings, no "I hope this helps!" closings, section bodies as plain prose (no markdown bullet-point sludge), hook-intro discipline, and a hard rule that the model NEVER invents statistics, studies, company names, quotes, or product specs you did not give it. You can do all of that in a long system prompt on a raw LLM — but you would rewrite it every time, and a teammate would not.
Yes — rule #1 in the system prompt is "NEVER INVENT FACTS." Concretely: when the model would want to cite a specific statistic, it either (a) uses a genuinely widely-known fact where it has high confidence ("Google processes over 8 billion searches per day"), (b) phrases it qualitatively ("many users report…", "a common pattern…"), or (c) leaves a bracketed placeholder for your real number ("[your starting price]", "[your 12-month retention rate]"). It will not invent a "73% of marketers said…" stat with a fake Forrester citation. This is the #1 problem with generic AI-written blog posts — fabricated sources get you flagged by fact-checkers, delisted by Google, and occasionally sued. The guard is not perfect (LLMs hallucinate) but the explicit rule + our editor-pass output schema cuts the invention rate way down, and we log the brief so you can audit.
Google's March 2024 Helpful Content Update specifically penalized thin AI-flavored prose, and every B2B blog on the internet now opens with "In today's fast-paced world…" The result: search engines, editors, and readers all recognize the fingerprint instantly. We hard-banned ~20 of the worst offenders across the full output: "revolutionary", "game-changing", "cutting-edge", "world-class", "industry-leading", "best-in-class", "unlock", "elevate", "supercharge", "seamless", "leverage", "unleash", "in today's fast-paced world", "in today's digital age", "at the end of the day", "when it comes to", "delve into", "plethora", "navigate the landscape" — plus "Let's dive in", "By the end of this post", "As we've seen", "I hope this helps!", and em-dash orgies. Removing them forces the model to write with actual specificity. You still get a human-adjacent voice, not a more-polished-AI voice.
Short (500-800 words, 3-4 sections, ~3 min read) is right for news posts, brief explainers, product-update announcements, and topics where the answer is genuinely small. Medium (900-1400 words, 4-6 sections, ~5-6 min) is the SEO default — it is the target length for most commercial and informational blog queries, balances depth against reader attention, and is what "how-to" searchers skim. Long (1500-2200 words, 5-8 sections, ~8-10 min) is for pillar content, comprehensive guides, and topics that reward depth — "complete guide to", "everything you need to know about". Long drafts take 75-130 seconds to generate (the longest step in our stack). The mistake most people make: using Long for topics that have 700 words to say. Padded long-form gets penalized.
Informative is the safe default — clear, neutral, expository, third person, short-to-medium sentences. Use it for how-to, explainer, news, glossary, enterprise content. Conversational is second person ("you"), contractions, warm and direct — like talking to a smart friend. Use it for DTC brand blogs, creator newsletters, lifestyle content, consumer SaaS. Professional is polished business voice — third person or formal "we", no contractions, no slang, authoritative. Use it for B2B, finance, legal, compliance, enterprise. Casual is loose and energetic with contractions everywhere, the occasional one-word sentence, slightly irreverent. Use it for food, travel, entertainment, creator blogs with personality. If in doubt, pick informative or conversational — they outperform the other two on both SEO ranking and reader retention in most categories.
No — but results are noticeably better if you do. Without an outline, the model makes up a 4-6 section structure based on title inference, which is competent but generic. With an outline (one section idea per line), it produces sections that map exactly to what you already decided the post should cover, which means a far less painful editing pass. The best workflow: start with our Essay Outline Generator to produce an outline tuned to your keyword and audience, paste it into the Outline field here, and generate the draft. The seam is frictionless — both tools use the same underlying model family, and the outline format lands directly in the outline field.
Comma-separated, primary keyword first, 1-4 total. The model will weave the primary keyword into finalTitle, metaTitle, metaDescription, the first H2, the intro, one section body, and the conclusion. Secondary keywords appear once each in section bodies or H2s. No keyword stuffing — if a keyword would sound forced, the rule says omit it rather than break voice. This is not a substitute for real keyword research — use our Keyword Research Generator to find the keywords first (it clusters them, tags intent, estimates difficulty), then paste your primary + 2-3 supporting terms here for the draft.
Within ±10% for short posts, ±15% for medium, ±20% for long. The model sums words in intro + section bodies + conclusion and reports it; reading time is word count ÷ 240 (standard silent-reading speed). We do not count keyFacts or pull quote (they are scannable UI elements, not prose). Word counts should be treated as ballpark — if you paste into WordPress, Ghost, or a Markdown editor, use that editor's count as ground truth. The estimate is mainly there so the draft feels right-sized at a glance before you commit to editing.
First 3 runs per day are free without signup — enough to draft a week's worth of posts if you space them out. Free accounts get 30 credits/month; each blog post draft costs 4 credits (full drafts are expensive to generate, so the per-run cost is higher than for lighter tools like FAQ or title generators). Pro accounts get 1,500 credits/month — enough for daily posts across multiple brands or a full editorial calendar.