More sites, scattered status
Multi-site teams switch between a dozen admin panels and track stale categories, stuck articles, and lagging sites by hand.
PrioritiesBulk setup, category inventory, concurrent generation, schedules, retries
AI Publish by YunAISEO
One console for multi-site AI writing,
scheduled publishing, and search submission
For SEO agencies, multi-site operators, and enterprise content teams: manage sites, categories, titles, AI settings, publishing jobs, and results in one place. Works with 18 CMS platforms; tasks are traceable, recoverable, and retry-safe.
- 18 CMSs and site builders
- Use your own AI API keys
- Traceable, recoverable, retry-safe tasks
- Retries never double-publish or double-charge
PRODUCTION RAILOne workflow, five recoverable stages
- Sites
- Titles
- Generate
- Publish
- Search submission
Connect your stack. Keep your sites.
WordPress / EmlogTypecho / PbootCMSGhost / ShopifyDuda / WixBaidu / Bing
Your bottlenecks
Multi-site content operations
usually break in three places
When teams juggle many sites, categories, and CMSs, the bottleneck is not writing — it is lost status, repeated work, and no audit trail.
More batches, messy reviews
Agencies stitch job scope, templates, pricing, and delivery records from spreadsheets across mixed client CMSs.
PrioritiesCMS support, job snapshots, status tracking, credit reconciliation, proof of delivery
Manual publishing breaks cadence
Enterprise teams need steady coverage of products and long-tail queries, but manual publishing makes brand, SEO, and cadence rules hard to enforce.
PrioritiesBrand consistency, internal links, images, publishing, search submission
Not for occasional one-off articles
Best for teams with existing sites, categories, ongoing publishing goals, and fewer manual handoffs.
Three steps
Connect sites → generate and review →
scheduled publishing and tracking
Run the full pipeline in three steps: connect existing sites, generate and review with your own AI key, then hand off to scheduled publishing and search submission. Every input, state, and result stays verifiable.
-
Connect your sites
Verify domains one by one or in bulk, read categories from any CMS, and store mappings, language, and timezone — no rebuild needed.
Output: ready sites and categories -
Generate and review
Batch-generate articles from your title inventory with your own AI key; review drafts before publishing, with dedupe and quality gates applied up front.
Output: approved article drafts -
scheduled publishing and tracking
Publish on category schedules, then auto-submit to Baidu and Bing; one failing channel never blocks the others, and retries never double-publish.
Output: article URLs and submission status
Job snapshots
Background task queue
Retry failed tasks
Duplicate protection and activity history
Product evidence
Real interface screenshots,
click to enlarge
What you see is not an “automation” promise — it is scope, settings, status, progress, failure reasons, and next actions. Click any screenshot to open the full-size image.
18 supported platforms
Keep existing sites,
connect them once
Self-hosted PHP CMSs connect through a root connector; hosted builders use official APIs. Categories sync on demand and the platform never changes your site structure automatically — both end up as one mapping of sites, categories, articles, and remote links.
01 / ROOT CONNECTOR
Upload one connector file14 PHP CMS root connectors
Verify ownership, sync categories, publish articles, process images, and return CMS IDs; requests allow only approved platform traffic.
02 / OFFICIAL API
No PHP upload requiredConnect four hosted platforms
Ghost
Admin API v6Sync tags, create or update articles, and assign author and content fields.
Shopify
Admin GraphQLSync blogs, publish articles to the selected blog, and preserve remote mappings.
Duda
Blog APIConnect Duda blogs with site IDs and authorization, then return each publishing result.
Wix
Blog API v3Connect the Wix app and author, sync categories, and create blog articles.
Verify sites and sync categories
Publish, update, and return URLs
Retry failures and track status
Encrypt credentials and verify ownership
Verified cases
Verifiable cases only,
no invented numbers
Every case documents the original workflow, site count, CMS types, scope, time saved, failure recovery, and limits. Ranking gains appear only as supporting evidence — rankings also depend on site authority, competition, and search policies.
Before and after
How sites, writing, and publishing worked before and after adoption — including what is still handled manually.
Scale and scope
Site counts, CMS types, category counts, and article batches; customers may stay anonymous.
Results and limits
Time saved, failure recovery, and job volume — plus the scenarios not covered and the limits that remain.
First public cases in preparation
If you would like to share your original workflow, site scale, and results (anonymously if preferred) as one of the first public case studies,contact us.
Security and reliability / activity log
See what happened at every step
Credits, jobs, publishing, and search submissions are logged separately. If something goes wrong, you can see where it stopped, what was affected, and what to do next. For key protection and data deletion, seeSecurity.
Save the settings for each task
Each task keeps the model, template, target, and price used at submission, so later edits do not change past tasks.
Continue after an interruption
Task progress is saved so work can continue after a short interruption.
Prevent duplicate work
Repeated clicks or system retries do not duplicate charges, generation, or publishing; important changes are recorded.
User-owned encrypted credentials
AI, Baidu, and Bing credentials stay encrypted and never appear in user views, admin pages, or standard logs.
Loading service statusYou can still browse this page. Sign in to review notices and service status before running jobs.
LimitsAI providers bill model usage; CMS and search availability depend on third parties, while the platform records, retries, and alerts.
No ranking promisesRankings still depend on content, authority, competition, and search engines; the platform focuses on consistent, measurable operations.Contact support
Plans and costs
Three plans,
Prices follow the interface language.
Simplified Chinese shows CNY prices; Traditional Chinese and English show USD prices. New purchases use the selected language price, while paid orders keep their original amount.
| Plan | Monthly | Annual (two months free) | Active sites | Successful articles/month | Concurrent AI tasks | Article overage / article | Active-site overage / month |
|---|---|---|---|---|---|---|---|
| Starter | $19 | $190 | 3 | 150 | 2 | $0.12 | $8 |
| Growth | $59 | $590 | 10 | 800 | 5 | $0.08 | $6 |
| Agency | $179 | $1,790 | 30 | 3,000 | 12 | $0.05 | $5 |
Larger scaleScale and Enterprise plans serve teams running 100–300+ sites with higher quotas and concurrency; view them in the app after signing up, or Contact supportrequest the full price list.
Price book2026-09-v2 CNY prices include tax; USD prices exclude local taxes. The trial fee counts toward your first month if you upgrade within 30 days.
Advanced-feature prices and included quotas
Included amounts reset monthly. Included usage is used first; only successful results above the allowance cost credits. Failed or canceled work and automatic retries do not cost credits.
| Advanced feature / how it is counted | Price above the included amount | Trial | Starter | Growth | Agency |
|---|---|---|---|---|---|
| Site analysis / each successful analysis | 10 points | 1 analysis | 3 analyses | 10 analyses | 30 analyses |
| Content opportunities / each successful batch (up to 100) | 5 points | 1 batch | 3 batches | 20 batches | 100 batches |
| AI title generation or rewrite / each successful title | 1 points | 20 titles | 0 titles | 300 titles | 2,000 titles |
| Content test / each successful test | 3 points | 2 tests | 0 tests | 10 tests | 50 tests |
| Deep SEO check / each successful article | 3 points | 5 articles | 0 articles | 100 articles | 800 articles |
| Smart rewrite or Humanize / each successful article | 5 points | 2 articles | 0 articles | 30 articles | 250 articles |
AI providers bill model usage
Use your own AI API key; provider token charges are not included in any platform plan.
Only successful results count
Failed and canceled work, plus automatic retries of the same result, are not counted twice. Successful articles use the plan article allowance; advanced features use credits.
Off by default, with a required cap
When included quota runs out, new paid actions stop by default. To continue, enable overage and set the most you can pay each month.
Points cover advanced features only
Credits do not pay plan fees or add to the plan article allowance. Site analysis, title generation, content tests, deep checks, and smart rewrites are charged in credits.
7-day trial$1.99, Includes one site and 20 successful articles; upgrade within 30 days to credit it toward month one.
Start 7-Day Trial · $1.99FAQ
What to know before you start
Plan limits, third-party costs, and publishing methods are shown before you submit a job.
Which CMSs and site builders are supported?
Root connectors support WordPress, Emlog, PbootCMS, Typecho, Z-BlogPHP, DedeCMS, EmpireCMS, PHPCMS V9, Joomla, Drupal, MODX, ProcessWire, MetInfo, and DESTOON; official APIs support Ghost, Shopify, Duda, and Wix.
Do I need to rebuild my site?
No. Upload a connector to self-hosted PHP CMSs or authorize the hosted platform API; AI Publish keeps your theme and frontend unchanged.
Who pays AI model fees?
You provide the AI provider key, and the provider bills model usage directly; platform plans cover operations and capacity.
Do plans have different features?
Core site, opportunity, title, generation, publishing, and search workflows remain available. Plans differ by active sites, successful articles, AI concurrency, and included advanced-feature quotas; published overage rates apply above them.
How are credits charged for advanced features?
Each plan uses its included site analyses, opportunity batches, AI titles, content tests, deep SEO checks, and smart rewrites first. Credits are charged only for successful results above those allowances; failed, canceled, and automatic retry results are not charged.
Is data deleted when a plan expires?
No. Expiry makes data read-only and stops new jobs and automatic publishing; renew within 30 days before retention rules may remove data.
How are duplicate generation and publishing prevented?
The platform recognizes the same task and result, so normal retries do not generate, publish, or charge twice.
Can published articles be submitted to search engines?
Yes. After CMS publishing, AI Publish can submit to Baidu and Bing asynchronously; submission stays separate and never guarantees indexing or ranking.
Do failed remote images block publishing?
No. The connector tries remote or local images in order; if all fail, it publishes the article and records a warning.
Start your workflow
Connect one site.
Publish one complete article.
Verify the site, sync categories, import a title, generate an article, and publish with every key state tracked.