Written by David Roberts
Reviewed by Lucas Walter
Updated
The top AI influencers in 2026 are Fireship, Vaibhav Sisinty, Jeremy Carrasco, Karen X and Rourke Sefton-Minns. This list ranks 25 AI influencers by median views across their recent posts, not by follower count. Each one was found by searching across hundreds of thousands of top performing posts in the AI niche.
Examples of what you can ask Klint:
Join 1,000+ marketers who use Klint to find creators in their niche. Start free
All 25, in rank order. Select a row to jump to the write-up.
| Rank | Creator | Followers | Median views | Engagement | Known for |
|---|---|---|---|---|---|
| 1 | Fireship @Fireship | 4.3M | 1.4M | 1.6% | Fast AI and coding news |
| 2 | Vaibhav Sisinty @vaibhavsisinty | 2.2M | 495K | 1.3% | AI news, anchor style |
| 3 | Jeremy Carrasco @jeremyfindsai | 516K | 230K | 4.6% | Spotting AI-made videos |
| 4 | Karen X @karenxcheng | 1.5M | 212K | 3.3% | Things invented with AI |
| 5 | Rourke Sefton-Minns @rourke | 977K | 211K | 3% | AI video and VFX tutorials |
| 6 | Varun Mayya @thevarunmayya | 1.3M | 200K | 2.7% | AI and tech news from India |
| 7 | Greg Isenberg @GregIsenberg | 727K | 158K | 1.9% | Startup ideas built on AI |
| 8 | Catharina Doria @cahdoria | 657K | 152K | 7.2% | AI ethics and literacy |
| 9 | Nick Saraev @nick_saraev | 705K | 138K | 4.4% | Claude Code and AI agents |
| 10 | Peter Steinberger @steipete | 603K | 133K | 0.6% | Building OpenClaw in public |
| 11 | Two Minute Papers @TwoMinutePapers | 1.8M | 131K | 3.5% | AI papers in five minutes |
| 12 | Wes Roth @WesRoth | 330K | 119K | 1.7% | AI lab news, daily |
| 13 | Maverick Maltin @mavgpt | 1.4M | 117K | 3.5% | ChatGPT prompts and codes |
| 14 | Jeff Su @JeffSu | 1.9M | 107K | 2.2% | AI for office work |
| 15 | AI Explained @aiexplained-official | 455K | 101K | 4% | Frontier AI, read closely |
| 16 | Will Francis @willfrancis24 | 262K | 94.3K | 4% | ChatGPT and Claude for marketers |
| 17 | Claire Zau @zauey.talks | 274K | 90.1K | 2.5% | AI news from an investor |
| 18 | Riley Brown @rileybrownai | 300K | 81.8K | 1.3% | AI agents for knowledge work |
| 19 | Derek Cheung @derekcheungsa | 446K | 80.6K | – | No-code AI automation |
| 20 | Cindy Zhu @cindiezhu | 236K | 71.5K | 5% | No-code AI tools |
| 21 | Nate Herk @nateherk | 1.1M | 71K | 1.6% | AI automation tutorials |
| 22 | Simon Willison @simonw | 238K | 68.1K | 1% | Language models, tested |
| 23 | bycloud @bycloudAI | 237K | 64.5K | 4% | AI research explained |
| 24 | Robert Miles @RobertMilesAI | 173K | 64.4K | 6% | AI safety research |
| 25 | Elena Nisonoff @elenanisonoff | 198K | 63.7K | 10.9% | AI news with a punchline |
The best AI influencers aren’t simply the accounts with the most followers. YouTube, Instagram, TikTok and X all show posts to people based on how the post performs individually, not just based on who follows the creator. As a result, a small or mid-sized account can reach more people than one several times its size.
That’s why we rank AI influencers by median views of their posts (average number of people a typical post reaches). As of October 2026, the five most-watched are Fireship, Vaibhav Sisinty, Jeremy Carrasco, Karen X and Rourke Sefton-Minns.
Grouped by what each creator mostly posts, from launch-day news to finished AI videos. The number beside each name is the creator’s overall rank.
Model launches, lab news, and leaks, covered within a day. Worth a call when an announcement has a date.
Fast-cut explainers of the week’s AI and open-source news, narrated over code, terminals, and animated diagrams. Each video takes one event that just happened and ends with a sponsor read.
30 sponsored posts in the last 90 days, in developer tools, AI coding and agent tools, and cloud hosting.
News-anchor Reels on AI launches, research, and AI in business, each opening on a bold claim over screen footage. Posted almost daily from India, in English.
6 sponsored posts in the last 90 days, in AI video tools, AI chat apps and smart glasses.
Daily explainers on the newest AI launches and viral demos, mixing straight-to-camera commentary with clips and screen captures. Recurring team and family series run between the news.
11 sponsored posts in the last 90 days, in AI video tools, online courses and tech events.
The day’s model releases, leaks, and safety news, read straight to camera over a screen-recorded backdrop. Titles lean on urgency and capitals.
10 sponsored posts in the last 90 days, in AI video tools, AI note-taking apps and developer tools.
One-minute explainers of AI launches, funding rounds, and research results, filmed to camera, sometimes from an airport or in a storm. A venture capital partner by day.
No sponsored posts found in the last 90 days.
One AI or tech story per video, cut fast and dated in the caption, with a dry joke on top. Made by a media company cofounder, several times a week.
No sponsored posts found in the last 90 days.
Papers, safety arguments, and AI-made fakes, explained with care. These audiences expect evidence, not hype.
Short debunks that take a viral clip and show whether AI made it, narrated over the footage itself. The account’s line is media literacy for the age of AI video.
No sponsored posts found in the last 90 days.
Talking-head arguments about who AI helps and who it harms, with a recurring “Is this AI or not?” series and research breakdowns. An AI ethicist who also runs a community and a podcast.
No sponsored posts found in the last 90 days.
One AI research paper or model release per video, explained in about five minutes. Title cards, the paper’s own figures, and a scripted voiceover, with no camera.
21 sponsored posts in the last 90 days, in GPU cloud and machine learning tools.
Long, scripted videos that test the week’s AI claims against the sources, with charts and quotes on screen and no face. The presenter also built a reasoning benchmark for language models.
No sponsored posts found in the last 90 days.
Narrated explainers of frontier AI research, from model architecture to what inference really costs, drawn over paper walkthroughs. One person’s channel, made under a pen name with an editor and an animator.
12 sponsored posts in the last 90 days, in cloud communications.
Animated and to-camera explainers of AI safety and alignment research, plus Shorts reacting to AI-risk news. Plain language on why a powerful AI could go wrong.
No sponsored posts found in the last 90 days.
One tool or prompt at a time, for people who use AI at work. Suits products anyone can try today.
Podcast episodes and screen-shared tutorials on businesses to build with new AI tools and agents. Made for founders, by the CEO of a design studio.
7 sponsored posts in the last 90 days, in AI models and cloud services.
Screen-recorded Reels with one ChatGPT or Claude trick each, gated by a keyword to comment for the full list. Made for people who use AI at work and do not code.
6 sponsored posts in the last 90 days, in AI assistants, AI app builders and marketing tools.
Polished tutorials on using AI tools for everyday office work, from files and résumés to memory and agents. A product marketer by day, filming screen walkthroughs with on-camera intros.
3 sponsored posts in the last 90 days, in AI note-taking apps, marketing software and cloud storage.
Calm explainers on getting more out of ChatGPT and Claude at work, many of them numbered lists of settings or prompts. Launch-day posts cover new features as they ship.
6 sponsored posts in the last 90 days, in AI voice tools, design tools and AI video tools.
Talk-and-screen Reels on one AI tool or prompt set each, usually with a keyword to comment for the guide. Made for people with no technical background, by a former digital marketer.
20 sponsored posts in the last 90 days, in AI video tools and AI voice tools.
Coding agents, automations, and apps built on screen. Developers and founders watch to decide what to build with.
Screen-recorded builds with Claude Code and AI agents for beginners, each opening on a keyword to comment for the free setup. Longer tutorials run on YouTube.
3 sponsored posts in the last 90 days, in AI video tools.
Text posts from the founder of OpenClaw, an open-source AI agent, on releases, benchmarks, and how teams use it. Reactions to model launches and agent design fill the rest.
1 sponsored post in the last 90 days, in developer tools.
Long tests of what new AI assistants and agents can do, from two-hour courses to builds made on camera. Made by a cofounder of an AI app builder who also sets up agents for businesses.
18 sponsored posts in the last 90 days, in AI assistants, marketing software and developer tools.
Step-by-step builds of AI agents and automations for people who do not code, screen-recorded from start to finish. An engineer and instructor who also sells a course on the same tools.
21 sponsored posts in the last 90 days, in AI video editing and web data tools.
Hands-on tests of coding agents and AI assistants, screen-recorded with the code and the agent’s output in view. Made by an ex-Goldman Sachs AI consultant who now runs an AI automation community of 450,000 members.
16 sponsored posts in the last 90 days, in sales tools, AI agents and AI note-taking apps.
Text posts that test each new language model on price, prompts, and agent behavior, plus release notes for Datasette and an open-source command-line tool for language models. A co-creator of the Django web framework.
No sponsored posts found in the last 90 days.
Video effects and home inventions made with AI tools. The result opens the post, so the tool is on screen.
Inventions built at home with AI tools, such as a bot that writes a printed morning newspaper, filmed first person with the result shown. A few posts a month, each one a finished build.
8 sponsored posts in the last 90 days, in AI coding tools, AI image tools and AI chatbots.
Cinematic tutorials that turn ordinary footage into effects shots with AI video and image tools, edited fast with captions on screen. Most posts name the tool they use.
25 sponsored posts in the last 90 days, in AI video and image tools and AI assistants.
We analyzed the 423 posts these 25 creators published on YouTube, Instagram, TikTok and X in the last 30 days. Of those, 56 beat their creator’s usual views by 3× or more.
Here are the hooks that are working, read from each post’s caption, title, or on-screen text.
“comment ‘GPT’ for 100 more codes like these 💡”
The caption or title promises something to use straight away: a set of ChatGPT codes, a free setup, a numbered list of use cases. Of these 13 breakouts, 8 also ask viewers to comment a keyword, and the creator sends the rest by direct message.
Cindy Zhu 1.2M views · 16.2× their usualFor an AI product, the comment is the part to plan for. Everyone who types the keyword has asked for the next step, and the message they get back can carry your link and your free trial.
“OpenAI’s ‘Alien Math’ Is Freaking People Out”
The title or first line gives a verdict on a model, a lab, or the craft of coding before it explains anything. These 11 breakouts come from 8 creators, and none supplies more than 2.
Wes Roth 685K views · 5.8× their usualMost of these verdicts are about someone else’s product. A launch partner gets the creator’s verdict, not the brand’s, so this is the pattern for a product that holds up under a hard test.
“New ChatGPT dots! Major new feature brings an always-on 24/7 AI assistant…”
The first line says something has just happened, usually a launch, with the product’s name in the opening words. These 8 breakouts come from 6 creators on 4 platforms.
Will Francis 541K views · 5.7× their usualA creator who posts within hours of an announcement needs to know about it before it happens. For a launch, an embargoed briefing and early access are worth more to these creators than a script.
Maverick Maltin: “Comment ‘Tricks’ and I’ll send you 50 more codes 👇” reached 1.8M views, 15.7× their usual.
Elena Nisonoff: “agent problems (9.26)” reached 848K views, 13.3× their usual.
Varun Mayya: “Ben Affleck, the director and actor, just revealed that he writes Python, understands convolutional neural networks, and trains models on GPUs by night.” reached 1.9M views, 9.6× their usual.
What the AI influencers on this list do at each size on YouTube, Instagram, TikTok and X. Use it to judge whether a creator’s numbers are strong for their tier.
| Tier | Followers | Median views | Engagement | Posts a week | Sponsored posts |
|---|---|---|---|---|---|
| Mid | 100K to 500K | 80.6K | 4% | 2.6 | 22% |
| Macro | 500K to 2M | 145K | 3.2% | 2.6 | 19% |
| Mega | Over 2M | 946K | 1.4% | 4.5 | 49% |
Medians across these 25 accounts, from each account’s recent posts. “Sponsored posts” is the share of posts in the 90 days to Oct 13, 2026 that carry a paid-partnership or paid-promotion label or a stated sponsor, or are part of a brand campaign.
If you market an AI product, most of these creators have already been pitched by tools like yours this quarter, so the useful questions are about fit and conflicts, not reach.
Read a creator’s sponsored posts from the last 90 days before you pitch. AI creators often run two or three tools from the same category a few weeks apart, and their viewers notice. If it matters to you, ask for a period with no competing sponsor in writing.
A coding agent needs a screen recording long enough to show it finishing a task, which points to YouTube. A consumer app can land in a short Reel that opens on the result. A developer tool with a free tier can work as a text post on X with a link to try it.
Many AI creators run a course, a paid community, an agency, or software of their own. Ask whether your product competes with theirs, and how they will mention it next to their own offer.
Titles in this niche run on strong verdicts, and a sponsored video will repeat yours. Keep the brief to what the product does today: the FTC has brought cases against companies that used AI hype to deceive customers.
Some creators here put the paid-partnership label on a TikTok and post the same video to Instagram without it. Ask for the label on every copy: the FTC counts a financial relationship with a brand as something the post must disclose.
We identified and ranked these creators by searching hundreds of thousands of top performing AI posts across YouTube, Instagram, TikTok and X with Klint.
We rank by median views across each creator’s recent posts, so one viral post can’t lift an account up the list. Creators from any country qualify if they post in English and most of their recent posts are about AI; company accounts, AI-generated characters, and executives promoting their own products are left out.
To get this list or a similar one for yourself, describe the creators you’re looking for in the Klint web app, or paste this prompt into any AI agent connected to Klint:
Find the top 25 AI influencers on YouTube, Instagram, TikTok and X: real people who post mainly about AI, ranked by median views across their last 30 posts or their last 90 days, whichever is lower. Show followers, engagement and sponsors.
Klint is trusted by 1,000+ marketers for AI-native influencer marketing.
As of October 2026 the top five AI influencers by median views are Fireship, Vaibhav Sisinty, Jeremy Carrasco, Karen X and Rourke Sefton-Minns.
Among the people on this list, Fireship has the largest following, with 4.3M subscribers on YouTube, and also the highest median views. Andrej Karpathy has a similar following on X but posted only a handful of times in the last 90 days, too few to be ranked here.
On this list, yes. Every creator is a real person who talks about AI. Accounts run by AI-generated characters, such as Lil Miquela, are virtual influencers, and they are not on this list.
Mostly YouTube and Instagram: 11 of the creators here are listed on YouTube and 10 on Instagram, with 2 on TikTok and 2 on X. A creator with a large following on two platforms appears once, under whichever one gives their posts the higher median.
By hand, search YouTube and TikTok for reviews of the tools your product competes with, and note who covered their launches and who sponsored those videos. A faster route is to ask Klint for it in plain words, for example "AI YouTubers who reviewed coding agents this month and have no developer-tool sponsor".
It depends on the platform, the format, and how much of the post is yours: a sponsor read inside a YouTube video costs less than a whole video about your product, and usage rights or a period with no competing sponsor add to the fee. Many AI tools also pay part of the deal through an affiliate link or a discount code.
Yes. Any creator can ask to be taken off Klint’s lists, and the request applies to every page they would otherwise appear on.
On this list and would rather not be? Write to privacy@tryklint.ai and Klint will take you off.
