How LiHelper works: from profile sync to posts at your best time
Three steps: sync your profile and analytics, get a draft in your voice, and have it scheduled or published at your best hour.
LiHelper's AI LinkedIn post generator is only useful if the posts sound like you and go out when people will see them. LiHelper solves both halves in three steps: it syncs your real profile and analytics from your own browser, drafts each post with your headline, skills and best-performing content as context, and schedules it at the hour your own history says works. On Creator you approve each post with one tap; on Pro it can publish itself through LinkedIn's official API. Nothing needs your password, and nothing publishes until you say so.
Sync your profile and analytics
Everything starts with real data. You connect LinkedIn with OAuth, which lets LiHelper publish posts and read the basic profile you approve. Then you install the free LiHelper Connector Chrome extension, link it with a token from Settings, and click "Sync everything" on LinkedIn. In six steps it reads your analytics overview, content analytics across every time range, per-post stats for the past year, your recent posts with their text and dates, your profile (headline, current role, About, the Pages you admin) and your full skills list, and sends only those values to your account.
Two things come out of that sync. The AI gets a picture of what you do and how you write, taken from your actual profile and posts rather than a form you filled in. And the scheduler gets the history it needs to learn when your audience engages. Neither step needs a password or a session cookie; the extension page lists every permission and exactly what each sync step reads. The Free plan includes the sync, so you can look at your numbers before you pay for anything.
AI drafting in your voice
Each day, Daily Autopilot picks a topic from your synced profile — your industry, skills and headline, plus any focus you typed in and, where relevant, current headlines in your field — and writes a post. The draft is conditioned on your own best-performing posts as examples, retrieved fresh each time, so the tone, length and structure follow what has already worked for you. As you publish more and gather engagement, the pool of examples grows and the drafts track you more closely. No model is trained on your content; it is used as context for the draft and nothing more.
The models run through OpenRouter and currently include DeepSeek, OpenAI and Google models; you never need your own API key. Each post also gets a designed image — a hook card, numbered explainer, comparison or chart matched to the content — paid for with your monthly visual credits (15 on Creator, 40 on Pro; 1 credit = one designed image or chart; a card set uses 2, a carousel 3). If you run out of credits you still get the post, without an image, plus a Generate image button on the draft. You can also write or ask for a post yourself in the composer, and the FAQ answers the obvious question of whether AI posts will sound like you. The AI LinkedIn post generator page goes through the inputs, the image step and the plan quotas in full.
Scheduled at your best time
"Your best time" is a specific calculation, not a rule of thumb. Until you have enough history, LiHelper uses a LinkedIn benchmark slot: 10:00 on your own wall clock, in the time zone set on your profile. Once the extension has synced your post stats, it re-buckets your own posts onto your local hours and scores each hour by average engagement per post (impressions divided by ten, plus likes, plus twice the comments) — the same score the best-time guide explains. The hour with the best average becomes your slot — but only if it has earned a non-zero score over at least five scored posts, has at least two posts in that hour, is compared against at least one other sampled hour, and matches or beats the average of the hours sampled. Those guards stop a small account from crowning an accidental 3 a.m. post its "best time" and then reinforcing it forever.
The slot keeps updating as new data arrives, and every post is stored as a UTC instant so it fires at the right local hour wherever the server runs. Posts you write yourself use the same slot, on every plan including Free; the LinkedIn post scheduler page covers the free plan, the six-hour spacing rule and approval versus auto-publishing. When the slot arrives, the post publishes through LinkedIn's official share API with your OAuth token; no browser is driven and no button is clicked for you. How that keeps the posting side low-risk is covered on the security page.
Review or auto-publish: Creator vs Pro
On Creator, Autopilot drafts and schedules; you approve each post with one tap before it goes out, so nothing reaches LinkedIn that you have not read. On Pro you can switch on auto-publishing and the scheduled post publishes itself at your best time — you review results instead of drafts. Auto-publishing is off by default on every plan and can be paused at any time from Settings. Pro also adds the engagement autopilot and the Easy Apply and job tools, which run in your own browser through the extension within daily caps.
| Step | Creator | Pro |
|---|---|---|
| Profile and analytics sync via the extension | Included | Included |
| AI drafts conditioned on your profile and best posts | Included | Included |
| Designed image for each post | Included | Included |
| Best-time slot from your own engagement data | Included | Included |
| Daily Autopilot: drafted and scheduled for you | Included | Included |
| You approve each post before it publishes | Included | Included |
| Auto-publishing: scheduled posts go out on their own | Not included | Included |
| Engagement autopilot (likes and comments within caps) | Not included | Included |
| Easy Apply assist, cover letters and résumé optimizer | Not included | Included |
Creator: $12/month, 30 AI posts and 15 visual credits. Pro: $29/month, 60 posts and 40 credits. Annual plans cost 10 months' price. Full detail on the pricing page.
What happens every day
Autopilot runs on a background tick and plans one post per user per day; a 20-hour cooldown means it never plans twice in the same day, and it skips a day rather than draft a bad post if your profile cannot be read; if only the image fails, the post still lands without one, with a Generate image button on the draft. On Creator that post appears in your Drafts and calendar waiting for approval; on Pro with auto-publishing on, it sits in the calendar and fires at your slot. After you post, the next sync brings back the new impressions and engagement, which feed both the example pool for future drafts and the best-time model. That loop — sync, draft, schedule, measure — is the whole product.
Beyond posting: engagement and job tools
Posting is the safe core. Pro adds two things that act inside your own LinkedIn session through the extension: engagement autopilot, which runs likes you approved and drafts comments for you to review, within server-set daily caps and a circuit breaker that pauses on block signals; and Easy Apply assist, which fills applications from your saved details and stops at the review step, with AI cover letters and a résumé optimizer alongside. Both are off until you arm them. We are direct about the risk of in-session automation on the security page and in the FAQ.
How it works FAQ
Will an AI LinkedIn post generator make my posts sound generic?
A prompt-only generator will. LiHelper conditions every draft on your synced headline, current role, skills and the posts of yours that performed best, and it retrieves fresh examples of your own writing as you publish more. No model is trained on your data; your content is used as context for each draft. You still read every draft before it goes out, and on Creator nothing publishes until you approve it.
What does "your best time to post" actually mean?
Until you have enough history, LiHelper schedules at a LinkedIn benchmark slot: 10:00 in your own time zone. Once the LiHelper Connector has synced your post stats, it buckets your own posts by the hour they went out, scores each hour by average engagement per post (impressions divided by ten, plus likes, plus twice the comments), and moves your slot to the hour that has genuinely performed best for you. It needs at least five scored posts and at least two hours with two or more posts before it trusts a personal hour, and it keeps updating.
Do I approve every post, or can it publish on its own?
On Creator, Daily Autopilot drafts a post, designs an image and schedules it; you approve with one tap before it goes out. On Pro you can switch on auto-publishing, so the scheduled post publishes itself through LinkedIn's official API at your best time. Nothing is ever posted until you turn that on, and you can pause it any time from Settings.
Can I use LiHelper without the Chrome extension?
Yes. Writing, scheduling and publishing posts work in any modern browser after you connect LinkedIn with OAuth. The extension is what syncs your analytics and profile, so without it drafts lean on what you tell LiHelper about yourself and the best-time model stays on the 10:00 benchmark until it has data. Install it when you want the numbers and the personal slot.
How many posts does Autopilot create?
About one a day. Daily Autopilot plans a single post per user per day (a 20-hour cooldown prevents two in one day), which fits the 30 AI posts a month on Creator. Pro doubles the quota to 60 posts and 40 visual credits, so you can add a second daily post from the composer. Posts you write yourself never count against the quota.
Every account starts with a 14-day Creator trial — no card needed. Want Pro? Start a 14-day Pro trial at checkout; you are not charged until it ends. Create your account, connect LinkedIn, install the extension and let the first draft land in your calendar. More on the team behind it on the about page.