SEO Consulting

Koala AI Review November 2026: From an AEO Consultant

A November 2026 Koala AI review from an AEO consultant: word allowances, autopilot, and which plan actually publishes the calendar.

Koala AI is the better fit when one paid plan has to hold two meters at once: a word allowance for articles, and a separate credit balance for images, links, research, and the agent, with publishing to a CMS you already run. It is a weak first buy when someone treats the headline about ChatGPT, Perplexity, and Claude as a stored prompt log. On 21 November 2026 the pricing cards showed Starter at $25 a month or $240 billed annually, Professional at $49 or $468, and Boost at $99 or $948. Growth through Scale III keep climbing, from $179 a month to $2,000. The free card is $0, and it is a one-time allowance, not a monthly plan.

Last updated - November 2026

I am Austin Heaton, an AEO consultant. The content systems I review are mostly for fintech. The rest are B2B SaaS and a few other industries when the brief looks the same. A writer login is useful when I can see the word cap, the model that multiplies it, and whether the post will land as a draft. It is not useful when a team will ship a generated fee, a license, or a partner, or when a sentence about getting cited is doing the work of a plan card.

The monthly prices, the annual totals, the yearly word and credit lines, the brand counts, the agent message caps, and the FAQ are from Koala's pricing page, read on 21 November 2026. The page was showing both a monthly figure and an annual figure on each paid card, with a control labeled Monthly, Annual, and Save 20% with annual. I did not complete a checkout, so I do not have a receipt. The same ladder, restated as monthly words and monthly credits, is Plans and credits explained, which that page dates 9 November 2026. Where the card's yearly total divided by twelve matches that article's monthly line, I say so. Where a model name does not match, I leave both names in the review.

What Autopilot includes, and the $49 and $99 lines on that page, are from Koala's SEO automation page, read the same day. The older Brand DNA explanation, which still disagrees with itself about Professional's agent cap, is Brand DNA Plans Explained. That post's schema says it was published and modified on 28 August 2026, and the byline names Connor Smith as founder. WordPress statuses are from How to publish an article directly to WordPress, also dated 9 November 2026 on the help page. The AI search claims are from Koala's AI search optimization page. The company address, email, and phone in the review are the fields in the Organization schema on the pricing page. I did not create an account. I did not generate an article. I did not connect a site. I did not run an AI Overview check.

Two comparisons already on this site are not being rewritten as a second invoice. Koala beside Outrank, from a 9 November read, is Koala AI vs Outrank. Koala beside Byword, from a 13 November read, is Byword vs Koala AI. Those pages describe cards that did not print a per-plan annual dollar in the way the 21 November cards do. I am citing today's cards. I am not pasting Outrank's or Byword's prices into this table. A term score with a monthly credit balance on another login is the NeuronWriter review. A content score plus a prompt log is the SurferSEO review.

Key Takeaways

  • On 21 November 2026 the pricing cards showed Free at $0 for 5,000 words, 25 credits, 1 brand, and 5 agent messages, once, with no card required. Starter was $25 a month or $240 a year, with 540,000 words and 6,000 credits on the card. Professional was $49 or $468, with 1,200,000 words and 12,000 credits. Boost was $99 or $948, with 3,000,000 words and 30,000 credits. The support article's monthly figures are those yearly lines divided by twelve: 45,000 words and 500 credits, 100,000 and 1,000, 250,000 and 2,500. I am using both pages. I am not inventing a third allowance.
  • Growth is $179 a month or $1,716 a year, for 6,000,000 words, 60,000 credits, and 3 brands. Elite is $350 or $3,360, for 12,000,000 words, 120,000 credits, and 7 brands. Advanced is $500 or $4,800, for 18,000,000 words, 150,000 credits, and 12 brands. Scale I is $750 or $7,200, for 30,000,000 words, 180,000 credits, and 20 brands. Scale II is $1,250 or $12,000, for 60,000,000 words, 240,000 credits, and 40 brands. Scale III is $2,000 or $19,200, for 120,000,000 words, 300,000 credits, and 75 brands. Extra brand slots are $29 a month each, up to 10, on any paid plan.
  • The card says annual billing saves 20 percent. Starter, Elite, Advanced, and the three Scale cards land on an exact 20 percent of twelve times the monthly price. Professional, Boost, and Growth land a dollar or two under that arithmetic. Twenty percent of $49 is $9.80, and $49 minus $9.80 is $39.20, not the $39 the annual line implies. I am not correcting the card. A buyer who wants the charge should read the billing screen.
  • Words and credits are different tanks. The pricing FAQ says articles spend words. Platform credits are shared across KoalaChat, KoalaImages, KoalaLinks, KoalaMagnets, the Brand DNA agent, API access, and MCP research. The support article says an agent message also costs 4 credits, on top of the message cap. Monthly plans bank unused words and credits for up to 3 months of the plan's allowance. Annual plans hand you the year on day one, and the support article says those words and credits last 12 months. Agent messages on an annual Starter seat pool to 300. Professional pools to 600. Boost has no message cap.
  • Autopilot, the setting that generates and publishes scheduled articles, is on Boost and above. Professional is planning and on-request drafts, with 50 agent messages a cycle on the pricing card, the support article, and the SEO automation page. The August Brand DNA post says 25 turns in one section and 50 turns in another. Starter is 25 messages a month on the card. The AI search page describes an agent that can check live Google AI Overviews. I did not find a prompt cap or a stored ChatGPT log. No page I read guarantees a citation.

What Is Koala AI in November 2026?

Koala AI, on the pricing page a buyer can hand to a writer, is one subscription that covers articles, images, internal links, an SEO agent, and an MCP server for an assistant the buyer already pays for. The hero line is simple pricing for growth on autopilot. The schema on that page calls it an AI SEO platform: KoalaWriter for articles, an agent that plans from Search Console data, and tools for images, internal links, and AI search. The same schema names Koala AI, the URL koala.sh, a Gainesville, Florida street address, the email hello@koala.sh, and a phone number. I am citing the schema. I did not call the number, and I did not confirm the office.

Want AI engines to recommend your company? Book a free 30-minute AI visibility review with Austin.

Book a Call

The product is not one button. KoalaWriter is the article generator. The word count on the card is the meter for that generator. KoalaImages, KoalaLinks, KoalaMagnets, and KoalaChat spend the other meter, platform credits. Brand DNA is the workspace: one brand holds the agent, a content calendar, a knowledge base, integrations, and writing defaults, which is how the August post describes it. The SEO automation page says the agent re-reads Search Console and Analytics, checks keyword volumes and competitor rankings, and puts striking-distance keywords and competitor gaps on a calendar with a keyword, an angle, and a date. KoalaWriter then drafts from that brief. That is the loop the marketing page is selling. The plan card is where the loop gets cut, because Autopilot is not on every card, and the agent message cap is not the same number on every card.

The free card is easy to brief and easy to over-brief. It is $0, one time, no credit card. The bullets are 5,000 words, 25 credits, 1 brand, and 5 agent messages. The FAQ says the same word and credit grant, and it says you can upgrade later. The card says the brand and the articles carry over. It also says you can try a demo without an account. I did not open the demo, so I cannot tell you whether those 5,000 words are available before a login or only after one. I am not going to call the free card a trial of Boost. Five agent messages is a preview. It is not the uncapped agent.

The job I can defend is narrower than the hero. Someone has to choose the query the business can actually serve. Someone has to catch a wrong fee in a paragraph the model drafted. Someone has to decide whether the WordPress status is draft or published. Koala's pricing page can explain a word cap, a credit cap, a brand cap, and a split between a writer that is on Starter and a calendar that publishes itself on Boost. It does not, on the pages I read, close the fee check by itself. "Get cited" is a goal on the AI search page. It is not a promise that an answer engine will mention the brand. I did not find that promise.

How Do the Plans Compare?

The first table is the entry ladder as I could read it on 21 November 2026. Yearly words and credits are the figures printed on the cards. The support article's monthly equivalents are the same numbers divided by twelve, and I put that equivalent in the cell so a budget that thinks in months can see it. Enterprise is not a column. Koala does not sell a custom column on this page. It sells a longer ladder.

CriterionFreeStarterProfessionalBoost
What the card says it is forA one-time allowance. No credit card. The card says you can also try the demo without an account.The writer, images, bulk mode, publishing, the API, and the MCP server, with a small agent cap.Everything in Starter, plus the editor score, Search Console and Analytics, internal linking, lead magnets, and Deep Research.Everything in Professional, plus the full Brand DNA experience, an agent with no message cap, and Content Calendar Autopilot.
Price I could read on 21 November 2026$0. One time.$25 a month, or $20 a month billed annually at $240. The 20 percent label is exact here: 25 times 12 is 300, and 80 percent of 300 is 240.$49 a month, or $39 a month billed at $468. Strict 20 percent of 49 is 39.20, not 39. Twelve times 39 is 468, which is the billed line.$99 a month, or $79 a month billed at $948. Strict 20 percent of 99 is 79.20. Twelve times 79 is 948.
Words and credits5,000 words and 25 credits, once. Not a yearly pool.540,000 words and 6,000 credits per year on the card. Support article: 45,000 words and 500 credits a month. 540,000 divided by 12 is 45,000. 6,000 divided by 12 is 500.1,200,000 words and 12,000 credits a year. Support: 100,000 words and 1,000 credits a month. The division is exact.3,000,000 words and 30,000 credits a year. Support: 250,000 words and 2,500 credits a month. The division is exact.
Brands and agent messages1 brand. 5 agent messages.1 brand. 25 messages a month. Annual pool on the support article: 300, which is 25 times 12.1 brand. 50 messages a month on the card, the support article, and the SEO automation page. Annual pool: 600.1 brand, described as the full Brand DNA experience. No message cap. The August post says the same: turns stay uncapped, and the included brand slot is 1.
Writer, editor, and researchThe card does not list bulk mode, the editor score, or Deep Research.Bulk writing. Real-time factual data. SEO optimization. AI images and YouTube embeds. Live Amazon data. The card names GPT-6 and Claude Sonnet 5.5. It does not print the word multiplier on the bullet.Article editor with AI chat edits and an SEO score. AI-powered editing. Deep Research. KoalaLinks for up to 50 sites. KoalaMagnets. Search Console and Analytics in one dashboard.The same writer lines, plus 2x faster bulk article creation. Autopilot is the new switch, not a larger editor.
Publishing and accessThe FAQ says free accounts cannot connect an MCP app. They can see the tool list.WordPress, Shopify, Webflow, Ghost, and webhooks. API access. MCP included. Limited file and image uploads. The support article says automatic WordPress categories need Professional or above.The same publishing list. Automatic categories are documented here. MCP still spends the same credits. There is no separate MCP fee.The same publishing list. Autopilot uses the saved WordPress defaults, on the help article's wording about articles that publish automatically.

If the word cap and the Autopilot switch are about to be two purchases, Book a Call and bring one page you would actually refresh. The page shows whether you need the writer, or a calendar that is allowed to publish.

The number I would circle is not the $25. It is the jump at Boost, and then the fact that Boost still has one brand. Starter to Professional adds the editor score, the Search Console dashboard, KoalaLinks, KoalaMagnets, Deep Research, a second step of agent messages, and about double the words and credits. It does not, on the card, add Autopilot. Boost is the first price that lists an uncapped agent and Content Calendar Autopilot. It is also the first price whose bulk line says 2x faster. A founder who hears "autopilot" on the hero and then buys Starter has bought the writer. The calendar that runs itself is a different card. I would not brief a team that every paid plan publishes on a schedule. The SEO automation page says each draft waits for review unless Autopilot is on, and it puts that switch on Boost.

Professional's agent cap is the cell that will start an argument if someone brings the August post into procurement. The pricing card says 50 messages a month. The support article, dated 9 November, says 50. The SEO automation page says 50 agent messages a cycle at $49. The August post's plan list says Professional is a lighter tier with a 25-turn allowance. The same post's subscriber section says the agent, which ran uncapped through the beta, now includes 50 turns per cycle, and that an annual plan pools 600. I am not averaging 25 and 50. The live pricing page and the November support article agree with each other, and with the second section of the August post. The first section of that post is the outlier. If a seat still shows 25, the account wins. This review has the pages, not the meter inside a login.

Starter's 20 percent is the clean cell. Twenty-five times twelve is $300. Eighty percent of $300 is $240, which is the billed annual line, and $240 divided by twelve is $20, which is the annual monthly figure on the card. Professional is not that clean. Forty-nine times twelve is $588. Twenty percent of $588 is $117.60. Five hundred eighty-eight minus $117.60 is $470.40, and the card bills $468. Thirty-nine times twelve is $468, so the annual total matches the $39 they printed, and the $39 is twenty cents under a strict 20 percent of $49. Boost is the same shape. Ninety-nine times twelve is $1,188. Twenty percent is $237.60. The remainder is $950.40. The card bills $948, which is seventy-nine times twelve. I would not tell a finance lead that every card is a clean 20 percent. I would tell them the billed annual numbers I could read, and I would tell them I did not flip a toggle that hid a different charge. The control's label and the card's own annual total were both on the page I fetched.

What the Higher Plans Add, and What They Do Not

Past Boost, the story on the cards is volume. The feature sentence is "everything in Boost," or, for Scale II and Scale III, "everything in Scale I." The things that change in public are words, credits, brands, and, on Scale I, a bulk-speed line. I did not find a new editor, a new CMS, or a prompt log that appears only at $750.

PlanMonthlyBilled annuallyWords per year on the cardCredits per year on the cardBrandsWhat else the card adds
Growth$179$1,716, shown as $143 a month. Strict 20 percent of 179 is 143.20.6,000,000. Divided by 12, that is the support article's 500,000 a month.60,000. Divided by 12, that is 5,000 a month.3Everything in Boost. The August post says Growth and up lift the agent turn cap. Boost already has no cap, so this is more brands, not a new switch.
Elite$350$3,360, shown as $280 a month. 350 times 0.8 is 280. Exact.12,000,000. Support: 1,000,000 a month.120,000. Support: 10,000 a month.7Everything in Boost. The card does not say everything in Growth. The brand count is the step.
Advanced$500$4,800, shown as $400 a month. Exact 20 percent.18,000,000. Support: 1,500,000 a month.150,000. Support: 12,500 a month.12Everything in Boost.
Scale I$750$7,200, shown as $600 a month. Exact.30,000,000. Support: 2,500,000 a month.180,000. Support: 15,000 a month.20Everything in Boost, plus 3x faster bulk article creation. The support article says Scale plans write up to 3 articles at once.
Scale II$1,250$12,000, shown as $1,000 a month. Exact.60,000,000. Support: 5,000,000 a month.240,000. Support: 20,000 a month.40Everything in Scale I, so the 3x bulk line comes along. I did not find a separate speed claim.
Scale III$2,000$19,200, shown as $1,600 a month. Exact.120,000,000. Support: 10,000,000 a month.300,000. Support: 25,000 a month.75Everything in Scale I. Same reading on the bulk line.

Credits do not scale in a straight line with words, and that is the procurement cell I would not tidy up. From Starter to Boost, credits and words both rise by the same factor if you compare the monthly equivalents: 500 to 1,000 to 2,500 credits, and 45,000 to 100,000 to 250,000 words. From Boost to Scale III, words go from 250,000 a month to 10,000,000, which is 40 times. Credits go from 2,500 to 25,000, which is 10 times. A team that buys Scale III because the word number looks like a content factory should also read the credit number. Images, link analysis, agent messages at 4 credits each, and MCP research all draw on that smaller tank. The support article says KoalaWriter articles only spend words. If the work is articles and almost nothing else, the word line is the one that matters. If the work is research through an assistant, the credit line runs out first. I did not spend either one.

Brands are a hard cap with a public add-on, which is clearer than a lot of writer tools. The FAQ says extra brand slots are $29 a month each, up to 10, on any paid plan, changed from the account page. Ten times $29 is $290 a month, on top of the plan. That add-on does not, on the sentence I read, raise the word allowance. A Starter seat with 10 extra brands would still be a 540,000-word year, spread across more workspaces, if the FAQ means what it says. I did not add a slot. The August post has a different grandfather rule: if you created more brands during the beta than the plan now includes, those brands remain functional, and the limit applies to creating new ones. I cannot see who is in that group. A new buyer should not brief the beta exception as their entitlement.

The refund and the rollover bank are the other two sentences I would put next to the annual total. The pricing FAQ offers a 15-day refund on a new plan if usage stays under 15,000 words and 100 credits, and it says most refund requests are processed within 24 hours. Fifteen thousand words is one third of a Starter month, and it is a small slice of a Scale month. A team that generates a large batch in week one is outside that window even if the calendar is still inside 15 days. The rollover bank, on monthly plans, keeps unused words and credits up to 3 months of the plan's allowance. They are spent after the monthly allowance runs out. They do not expire while the plan is active. The bank is removed if the plan ends, including a failed payment, and lifetime-deal credits never go into it. Annual plans are a different clock. The FAQ says you get the whole year up front and the credits last 12 months. Switching away from an annual plan does not remove those credits. Buying a second annual subscription keeps the first pool on its original expiration. I did not change a plan, so the checkout warning the support article mentions, the one that tells you if a smaller bank will drop the overflow, is a sentence I am citing, not a dialog I clicked.

Lifetime deal holders get a separate discount the support article is willing to print and the pricing cards are not. The article says 30 percent off any plan at checkout, the lifetime allowances stack on top of the subscription, and sale-priced annual plans skip the discount and come with 30 percent extra words and credits instead. The August post says the same 30 percent, applied automatically. I did not find a lifetime price on the pricing page I opened. I am not going to invent one. If a seat is a lifetime deal, the invoice is the deal. This review cannot see that invoice.

Words, Credits, and the Model That Changes the Bill

The footnote on the pricing page is the sentence that makes the big word numbers smaller. It says the word count is based on GPT-6 Luna. GPT-6.1 Sol and Claude Sonnet 5.5 count words at 2x. Claude Opus 5.5 counts at 3x. A 2,000-word article on Luna is 2,000 words of the allowance. The same length on Sonnet 5.5, if the footnote is the rule, is 4,000. On Opus 5.5 it is 6,000. Starter's 45,000 words a month is 22.5 articles of 2,000 words on Luna, 11.25 articles on the double-rate models, and 7.5 articles on the triple-rate model, before any extra usage the support article warns about. I am showing the arithmetic. I did not generate the article, and the support article says usage can run higher than the length for other reasons. I am not going to guess those reasons.

The support article, dated 9 November 2026, does not use the same Sonnet name. It says the count is based on GPT-6 Luna, that GPT-6.1 Sol and Claude Sonnet 5 count words at 2x, and that Claude Opus 5.5 counts at 3x. Sonnet 5 and Sonnet 5.5 are not the same string. The Starter card's own bullet says GPT-6 and Claude Sonnet 5.5, which matches the pricing footnote and not the support article's "Sonnet 5." I am not going to decide that the help center omitted ".5" by accident. Both pages are Koala's. A security review that needs the model name should read the menu in the product. I did not open the menu. Opus at 3x appears on both pages, so that part is the one I would repeat without a footnote.

Credits are a rate card, and the pricing FAQ is specific enough to brief. A raw HTML fetch is 0.1 credits. A rendered page or a screenshot is 0.25. A live Google search is 0.5. Keyword and competitor lookups are 1. An article edit is 1. Backlink lookups are 2. Keyword volume is 3. Historical traffic is 4. Keyword ideas and variations are 1 credit per seed keyword. The FAQ also says one credit covers ten raw fetches, four rendered pages, or two searches. Ten times 0.1 is 1. Four times 0.25 is 1. Two times 0.5 is 1. Those three examples match the unit prices. A single question may call several tools, so a chat that feels like one prompt can be a stack of those rates. Failed lookups and incomplete fixed-price research calls cost nothing. A successful search with no matches still uses credits. I did not run a search.

The agent sits on both meters, which is the part a slide will get wrong. The card caps messages: 5, 25, 50, or none. The support article says a Brand DNA agent message costs 4 credits. Fifty messages on Professional, at 4 credits each, is 200 credits, out of a 1,000-credit month. That is a fifth of the credit pool before anyone generates an image or asks for a backlink lookup. Boost removes the message cap and does not remove the 4-credit line. An uncapped agent can still empty the credit tank. The FAQ says reading Brand DNA, the knowledge base, the calendar, and connected Search Console and Analytics costs nothing, and so does checking an article's status or reading a finished article. Research that reaches outside data, and article edits, spend credits. Generating an article through MCP uses the word allowance, not a credit, except that an edit is 1 credit. The assistant you connect is billed by its own provider. Koala's FAQ says that in one sentence, and it is the sentence I would keep when someone asks whether the MCP price includes Claude or ChatGPT. It does not. The MCP is included on every paid plan. Free accounts can see the tool list and cannot connect an app. There is no daily research cap. There are rate limits. I did not connect an assistant.

Monthly and annual do not refill the same way, and the support article is clearer than the hero. A monthly plan refills words and credits at renewal, and unused amounts move to the bank, capped at 3 months of allowance. An annual plan gives all 12 months on the day you buy. Agent messages follow that pattern on Starter and Professional: 300 and 600 for the year, usable any time, which is the support article's line and also the August post's line for Professional. I would not describe an annual Starter seat as "25 messages a month" if the pool can be spent in January. The card prints the monthly cap. The support article prints the annual pool. Both can be true if the product enforces the pool and also shows a monthly pace. I did not see the meter.

What Autopilot Publishes, and What It Holds

The SEO automation page draws a line the pricing hero blurs. Research and planning can run on a cadence: the agent re-reads Search Console and Analytics, checks live volumes and competitor rankings, and drops picks onto a calendar. Writing is KoalaWriter, using Brand DNA, with Deep Research pulling sources. Links and schema are described as KoalaWriter linking the new article as it writes, and KoalaLinks keeping older pages linked and adding entity schema. Publishing goes to WordPress, Shopify, Webflow, or Ghost. Anything else uses webhooks or the API, which the page says are on paid plans. The part that waits is the calendar. You approve or edit topics before anything is written. Each draft waits for review unless Content Calendar Autopilot is on. Autopilot publishes approved items on schedule and pauses when you switch it off. That is a setting, not a personality. I did not turn it on.

The plan gate matches across three pages, with one old sentence still sitting in the August post's plan list. The SEO automation page says Brand DNA runs on Professional at $49 with 50 agent messages, which covers planning and on-request drafting, and Boost at $99 includes the full experience: a complete knowledge base, an uncapped agent, and Autopilot. The pricing card for Boost says Content Calendar Autopilot. Starter's card does not. The August post says planning the calendar is on every plan that includes a brand, and autopilot, which generates and publishes each scheduled article, starts at Boost. It also says Starter carries 25 agent turns and that Autopilot starts at Boost. I am willing to brief that. I am not willing to brief the same post's "25-turn" line for Professional, because the rest of Koala's current pages say 50.

WordPress is the publish path I can describe without guessing a status. The help article, dated 9 November 2026, says you open a finished article, click Create Draft, and pick the site if more than one is connected. The options it names are the author, categories, post type, and post status. Statuses are Draft, Published, Pending Review, or Private. Automatic categories, where the model picks them, need Professional or above, and you can limit the pick to selected categories, with a default of 1 to 3. Plugin connections can set the author and can target a post, a page, or a custom post type. There is a beta switch to save the article as Gutenberg blocks. If you save the choices as the default, articles that publish automatically use them. If the status is Published, the article says the post is already live. If WordPress returns an error, Koala shows the message. I did not connect a site, so I have the help article and not a live URL. A team that wants a human in the path should set Draft and leave Autopilot off. A team that sets Published, and turns Autopilot on, has chosen a different product behavior. Both behaviors are documented. Neither one knows whether the fee in the draft matches the fee on your pricing page.

The pricing page's feature list adds objects the plan cards only hint at. Automatic SEO optimization means KoalaWriter looks at pages ranking for the keyword and covers entities and terms they share. Deep Research, from Professional, gathers sources before the article is written, with citations when you want them. Internal linking is described as indexing the site and placing links, and the page says over 25,000,000 links have been placed. That figure is their sentence. I did not audit it. AI images are named as GPT Image 2.6 and Nano Banana 2, embedded automatically or generated in KoalaImages. Amazon affiliate articles use live Amazon data, including reviews and product details, and that bullet is on the Starter card. The editor that grades a draft, with an SEO score against ranking pages, chat edits, and one-click Polish, is from Professional. I did not open a score, so I will not tell you the scale, the color, or whether a high score can be reached by stuffing terms. The page does not print a formula.

Sharing and cancellation are looser than a seat-priced product, and I would not confuse that with a team workflow. The FAQ says you may share the account within your organization and with people working for you. It does not print a seat count or a per-user price. Cancel is a button on the account page. Payment methods listed are credit cards, Apple Pay, Google Pay, US and Canadian bank transfers, WeChat Pay, Alipay, and Cash App Pay via Stripe, plus PayPal or Wise if you contact them for an annual plan. The FAQ says Stripe hosts the card data and Koala does not store it. I did not check out. Ownership of the output is also in the FAQ: you keep rights to what you put in, and Koala assigns you rights to the AI-generated output, with a pointer to the terms. I did not read the terms line by line. I am citing the FAQ's summary, not giving a legal opinion.

The testimonials on the pricing page are not a result I measured. The page says trusted by over 20,000 paid creators. It prints quotes, including a claim about sessions and revenue on a new site. I did not open the sites, and I am not using those figures as outcomes. A quote on a pricing page is a quote on a pricing page. It does not tell you whether the article that shipped had the right fee in it.

The AI Search Check, and the Log It Is Not

This is the section that decides whether Koala belongs in an AI visibility stack or only in a writing stack. The AI search page says the job is getting cited in Google AI Overviews, ChatGPT, Perplexity, and Claude, and that Brand DNA's agent tells you what to change. The checks it actually names are more specific than that list. An AI Overview audit checks keywords against live Google AI Overviews, reports whether your domain is among the cited sources, and compares you with competitors on those queries. Other lines are extraction-friendly rewrites, brand entity gaps, schema upgrades through KoalaLinks, and briefs that pull Brand DNA details into the draft. The sample questions are about pages that rank and do not show up in AI Overviews, FAQ schema, named entities, a URL audit, and questions to add. I did not ask any of them.

What I did not find is a prompt allowance, a daily refresh, a stored answer, or a sentiment score. The page does not say the agent runs a fixed list of questions through ChatGPT every morning and keeps the text. The introduction names ChatGPT, Perplexity, and Claude. The audit procedure it describes is Google AI Overviews. I am not going to promote the introduction into an entitlement. A recommendation to add an FAQ block is not a log of what ChatGPT said yesterday. If the buyer needs that log, this is the wrong invoice, and the neighboring reviews on this site are the ones that meter prompts. Koala's page also says answer engine optimization and generative engine optimization are two names for the same job. I am citing their wording. I am not using it to claim the product tracks every engine those names get attached to in a sales deck.

Schema and internal links are on-site work. They can make a page easier to quote. They do not place a paragraph on a publication you do not own. The SEO automation page says KoalaLinks adds links and entity schema on the site you connected. The AI search page says the agent can recommend content that fills entity gaps. Guidance is a change to your pages. It is not a byline on someone else's site.

Publishing from Koala lands on a CMS you connect. For a mention on someone else's publication, Offsite Posts writes and places industry articles that name the brand, and the site says those posts, mentions, and links are guaranteed while rankings, traffic, and mentions by ChatGPT, Gemini, Perplexity, or Google AI Overviews are not.

On a live account, Austin Heaton still reads the fee, the license, and the partner name before a generated paragraph ships. The SEO score, whatever scale it uses, measures coverage against ranking pages. It does not know the fee schedule. Autopilot will not do that read if the source page is wrong, and on Starter and Professional the Autopilot switch is not in the price. Deep Research can attach citations. A citation inside a draft is a link the model included. It is not proof the linked page says what the sentence says. I would open the link.

Who Koala AI Fits, and Who Should Skip It

Pick from the meter you can explain, the publish status you are willing to leave on, and whether a person will read the draft.

  • You want a printed price, a word allowance, and a draft you still approve. Starter, if 45,000 words a month on Luna, 500 credits, 25 agent messages, and 1 brand can hold the work, and if WordPress, Shopify, Webflow, or Ghost is the destination. Professional is the step that adds the editor score, the Search Console dashboard, Deep Research, and 50 agent messages. The engines you can name on those cards are writing models, not a tracker.
  • You need the calendar to generate and publish without opening every draft. Boost is the first card that lists Autopilot and an uncapped agent. It is still one brand. Read the WordPress status before you assume "publish" means draft. The help article says Published means the post is already live.
  • You need more than one brand in the price you can explain. Growth is 3 brands. Elite is 7. Advanced is 12. Scale I, II, and III are 20, 40, and 75. Extra slots at $29 are a different lever, capped at 10, and they do not, on the FAQ, add words. The Outrank comparison is the neighboring invoice if a fixed article count is the actual requirement. I am not rebuilding that price here.
  • You wanted the annual dollar and the monthly dollar on the same card. The 21 November cards print both. Earlier comparisons on this site, from 9 November and 13 November, were written when that annual dollar was easier to find on the support article than on the card. Use today's card for the billed total. Use the support article for the monthly word equivalent, which still divides cleanly.
  • You need a stored answer from ChatGPT or Perplexity. I did not find that product on these pages. The AI search page is an agent check, centered on Google AI Overviews, plus rewrite advice. The NeuronWriter review and the SurferSEO review are the logs to read if the objection is the tracker rather than the word cap.
  • You need the draft checked against a fee schedule. Nothing I read does that. The score is terms and structure against ranking pages. The agent can pull Brand DNA into a brief. Brand DNA is only as accurate as the site it read. On a fee page, a person still has to look.

A tool subscription and an engagement are different bills. This page is only the tool. The neighboring writing tools, if the objection is the editor rather than Autopilot, stay on the reviews already linked above.

What I Could Not Confirm

  • A generated article, a credit balance, or a WordPress post. I read the public pages. I did not create an account, and I did not start the demo.
  • Whether the billing screen charges the monthly rate, the annual total, or a tax line the FAQ does not print. I did not check out.
  • Which Sonnet name the product menu uses. The pricing footnote and the Starter card say Claude Sonnet 5.5. The 9 November support article says Claude Sonnet 5 for the double word count.
  • Whether Professional's agent meter shows 25 or 50. The August post says both, in different sections. The pricing card, the support article, and the SEO automation page say 50.
  • The SEO score's scale, and whether Polish spends words, credits, or neither. The pricing page names the score. It does not print a formula.
  • What the 25,000,000 internal-link line counts, and whether the 20,000 paid-creator line is accounts, seats, or something else. They are on the page. I did not audit them.
  • Whether an AI Overview audit stores the answer or only reports a yes or no for the domain. The feature page describes the check. It does not describe a prompt history.
  • A lifetime price, a per-seat price, and the price of a credit top-up. The FAQ describes sharing and the rollover bank. It does not sell a credit pack with a number I could put in the table. Extra brands at $29 are the add-on I could read.

Frequently Asked Questions

Who is Koala AI the better fit for?

Koala AI is the better fit when one subscription has to draft articles against a word allowance, keep a separate credit balance for images and research, and, on the plan that includes it, publish a calendar to a CMS you connect. It is a weaker fit when the team needs a stored prompt log across ChatGPT and Perplexity in the price they can explain, because the AI search page I read describes an agent check against Google AI Overviews, not a daily prompt allowance. The plan split is in this article. The prices are in the pricing table. I did not create an account.

Does every plan publish the calendar on its own?

No. The SEO automation page says you approve or edit topics before anything is written, and each draft waits unless you turn on Content Calendar Autopilot. That same page, the pricing card, and the Brand DNA post put Autopilot on Boost and above. Professional is described as planning and on-request drafting, with a message cap. Starter includes the writer and a smaller agent allowance, and the Brand DNA post says Autopilot starts at Boost. I did not turn Autopilot on.

Are words and platform credits the same allowance?

No. The pricing FAQ says article generation uses the word allowance, and platform credits are shared across chat, images, internal links, lead magnets, the Brand DNA agent, API access, and MCP research. The support article updated in November says the same split, and it says a Brand DNA agent message also spends credits. A monthly plan moves unused words and credits into a rollover bank. An annual plan gives the year up front. I did not spend either meter.

Which writing model changes how many words an article costs?

The pricing page footnote says the printed word count assumes GPT-6 Luna, that GPT-6.1 Sol and Claude Sonnet 5.5 count at twice that, and that Claude Opus 5.5 counts at three times. The support article updated in November names Claude Sonnet 5, not Claude Sonnet 5.5, for the double count, and it still names Claude Opus 5.5 for the triple count. I am not merging those model names. The Starter card lists GPT-6 and Claude Sonnet 5.5 without printing the multiplier on the bullet. I did not generate an article.

Does Koala AI track citations in ChatGPT the way a visibility product does?

Not on the pages I read. The AI search page says the agent can check keywords against live Google AI Overviews and report whether your domain is among the cited sources. The same page's introduction names ChatGPT, Perplexity, and Claude as places a citation can appear. I did not find a prompt cap, a daily refresh, or a stored answer log. A recommendation to change a page is not a log of what an assistant said yesterday. I did not run the audit.

What happens when an article is sent to WordPress?

The help article updated in November says a connected site can receive the article as a draft, a published post, pending review, or private. Automatic category selection needs Professional or above. The SEO automation page also names Shopify, Webflow, and Ghost, plus webhooks and the API on paid plans. If the status you pick is published, that help article says the post is already live. I did not connect a site.

Where do I read the current prices?

The figures I could verify are in the pricing table in this article, with the page each figure came from. The pricing cards show a monthly price and an annual total, and they say annual billing saves a percentage. A support article updated in November prints the same monthly prices and the same annual totals. A Brand DNA post from August prints a different agent allowance in one section than it prints in another. If a contract is already signed, the contract is the price.

Who should skip Koala AI?

Skip it when the job is a citation guarantee, because no page I read promises that an answer engine will mention the brand. Skip the plans below Boost when the requirement is a calendar that publishes without a person opening each draft, and read the Autopilot line before you assume every card includes it. Skip unattended drafts on any page that states a fee. The writer does not know the fee schedule. Skip it when the team will not read the draft, and this login is the one that gets ignored.

Read next:

If the word cap, the Autopilot switch, and the person who reads the fee are three different conversations, Book a Call and we will put the page, the plan, and the publish status on one list before anyone pays for a calendar the card does not include.

BECOME THE COMPANY AI RECOMMENDS.

Your next qualified lead starts with an answer.

Get more citations. Compete for the prompts that matter. Become the top choice in your category.

Book a Call