OpenAI said at DevDay on 29 September 2026 that ChatGPT has 1.2 billion weekly users. It is a company-reported reach number, not a measure of how many people click through to your site. For your traffic, the useful step is not to admire the figure but to measure what ChatGPT already sends you and whether your brand appears in its answers.
As of 30 September 2026. User figures are OpenAI-reported and not independently audited.
Key Takeaways
- The 1.2B weekly figure comes from OpenAI's own DevDay recap, where it refers to a "collective 1.2B weekly users."
- OpenAI did not publish a free versus paid split or growth rate in the material I read, so you cannot infer how many use search-like features.
- Reach is not referral traffic. AI answers often resolve the question without a click.
- Measure your own ChatGPT referrals and citations before changing strategy.
- Treat every platform number, including Google's, as vendor-reported.
What OpenAI actually said
The wording in OpenAI's DevDay 2026 recap is that OpenAI is "opening up ChatGPT as a shared surface where humans and agents can collaborate and where developers can directly launch new native experiences to our collective 1.2B weekly users."
That sentence is a pitch to developers. The number is there to say: build here, the audience is big. The Decoder reported the same claim alongside separate figures of over 35 million weekly ChatGPT Work and Codex users and 2.5 million businesses on OpenAI products.
What the number does not tell you
An analyst write-up at FourWeekMBA points out what OpenAI did not publish: prior figures for comparison, growth rates, or a free versus paid breakdown. That is analysis, not fact, but the gaps are real when you read the recap.
So three things remain unknown from the sources I read:
- How many of those weekly users ask questions where a brand website could be cited.
- How many click a citation when shown.
- How the count is de-duplicated across apps, devices and surfaces.
Reach is not traffic
A billion people using a product weekly says nothing about outbound clicks. Think of it like a giant newsstand: enormous foot traffic, but most people read the headline and move on.
My working rule as a content strategist: platform reach numbers describe the platform's opportunity, not yours. Your opportunity is the slice of prompts where your topic appears, multiplied by how often you are cited, multiplied by click behaviour. All three are measurable for your own site.
How this compares to Google's numbers
Google has reported its own scale: AI Overviews above 2.5 billion monthly users and AI Mode above 1 billion monthly users, per Google, as noted in my brief for this batch. Two cautions:
- Those are monthly numbers; OpenAI's is weekly. Do not stack them against each other as if the metrics match.
- Both are self-reported.
The practical reading: people are meeting answers inside AI surfaces at very large scale on both sides. Neither number tells you your click share.
Step 1: Measure the ChatGPT traffic you already get
You do not need a paid tool for a first look.
- In Google Analytics 4, open Reports, then Acquisition, then Traffic acquisition, and set the dimension to Session source / medium.
- Look for sources such as chatgpt.com or chat.openai.com. Referrer naming has changed over time, so search partial matches.
- Filter landing pages for those sessions and record: sessions, engaged sessions, conversions.
- Compare against organic Google for the same pages. Note that some AI-assisted visits arrive with no referrer and show as direct, so your number is a floor, not a ceiling.
- Repeat monthly and chart the trend, not a single month.
Do this before reading another opinion about AI traffic.
Step 2: Check whether you are cited at all
Traffic is a lagging signal. Citation presence comes first.
- Write 20 to 30 real customer questions in your niche, phrased how people talk.
- Run each in ChatGPT (logged out or a clean session where possible) and record: were you mentioned, cited with a link, or absent? Who was cited instead?
- Repeat on a schedule; answers vary run to run, so treat a single result as one sample.
- Log it in a sheet with date and prompt. It is manual and dull, and it tells you more than a dashboard subscription.
Step 3: Make pages worth citing
I keep this unglamorous and specific:
- Answer the question in the first two sentences of the page.
- Add primary-source links and dated facts, so a model or a person can verify.
- Use clear headings that match how questions are phrased.
- Publish original detail others lack: your own numbers, screenshots, process steps.
- Keep facts current; stale pages lose trust.
None of this is a ChatGPT trick. It is the same craft that helps in Google, which is why I do not chase platform-specific hacks.
What the DevDay features change
The announcement bundle included Sign in with ChatGPT with 16 partners as reported, plugin extensions, and an OpenAI Marketplace. Each is a possible new surface. They matter for product businesses building inside ChatGPT far more than for publishers waiting on citations. I have not tested any of them yet, so I will not claim results.
A simple decision table
| Your situation | What the 1.2B claim should change |
|---|---|
| Publisher relying on ad revenue | Track referrals monthly; do not assume growth. |
| B2B SaaS or services | Check citations for buying-intent questions. |
| Local business | Little; local intent still runs through maps and search. |
| Edtech or course seller | Test question-style pages and comparison content. |
| App builder | Read the developer platform terms; that is the audience the number targets. |
Red flags when someone sells you this number
- "1.2 billion users means huge free traffic for you." It does not follow.
- Growth projections built on the figure without your own referral data.
- Any tool promising guaranteed ChatGPT rankings. Answers vary and no vendor controls them.
A monthly reporting template
I keep a one-page AI visibility block in every monthly report, and it takes about thirty minutes:
- ChatGPT referral sessions, engaged sessions and conversions, versus last month.
- Citation sample: how many of the 20 to 30 tracked prompts mentioned or linked the brand.
- Top three pages cited or visited, and what they have in common (format, length, freshness).
- One action for next month, such as refreshing a stale page or adding a primary-source table.
If referrals are tiny, say so plainly. A small honest number is more useful to a client than a hopeful projection built on a platform headline.
What I could not verify
- Any independent audit of the 1.2B figure.
- The share of weekly users using search-style features.
- Actual click-through rates from ChatGPT citations, which I did not find in a reliable primary source.
FAQ
How many people use ChatGPT weekly in 2026?
OpenAI says 1.2 billion, per its DevDay recap on 29 September 2026. It is a company-reported number and I found no independent audit.
Does 1.2 billion weekly users mean more traffic to my website?
Not by itself. Reach counts people using the product, while traffic requires citations that people click. Check your analytics for chatgpt.com referrals to see your real number.
Is 1.2B weekly comparable to Google's user numbers?
Not directly. Google's AI Overviews and AI Mode figures are reported as monthly, OpenAI's as weekly, and both are self-reported.
How do I find ChatGPT traffic in Google Analytics 4?
Open Traffic acquisition, set the dimension to Session source / medium, and look for chatgpt.com or similar. Some visits arrive without referrer data and land in direct, so treat the number as a minimum.
Can I pay to appear in ChatGPT answers?
The brief for this batch notes ChatGPT ads run only on Free and Go tiers in limited markets, and ads are separate from organic citations. I would not treat any vendor's promise of guaranteed answer placement as credible.
What else did OpenAI say alongside the number?
The Decoder reported over 35 million weekly ChatGPT Work and Codex users and 2.5 million businesses on OpenAI products. These are also company-reported.
Should I change my SEO strategy because of this?
Not because of the headline number. Keep building pages that answer questions clearly with verifiable sources, and add citation tracking to your monthly reporting.
How often should I check whether ChatGPT cites me?
Monthly is enough for most sites, with a fixed prompt set so results are comparable. Answers vary run to run, so look at trends across many prompts.
Where can I read the original claim?
In OpenAI's DevDay 2026 recap at openai.com. The 1.2B line is in the section on opening ChatGPT to developers.
Talk to Younus
I am Younus Fardeen, an India-based organic growth, SEO and content strategist with 4+ years of marketing experience across edtech and startups. If you want help measuring what AI platforms really send your site and building content worth citing, see my work and reach me through the contact form at younusfardeen.in.