Aug 27, 2025 · 26m · a16z
The Top 100 Most Used AI Apps in 2025
0:00 / 0:00
▶ Watch on YouTube →
gold bands on the timeline = statements, start to end. Hover to read, click to jump. CC turns on captions
a16z partners Justine Moore and Olivia Moore present findings from the fifth biannual Top 100 Consumer Gen AI Apps report, tracking web and mobile active usage across major global AI products. They analyze key industry shifts, including ecosystem stabilization, Google's multi-product traffic, Chinese AI export strategies, vibe coding monetization, and defensible moats for enduring AI applications.
How this conversation actually went
Every chapter scored 0–10 on four independent dynamics. Hover any point for the reasoning behind the score. How this is scored →
The host as informed peer 4.6 Guest teaching 3.6 Guest disagreement 0.0 The host pushing back 0.0
speaking balance: gold is the host, purple is the guest (3 minute bins)
Sharpest disagreement ▶ 5:03 Narrative violation on NotebookLM retention
In a highly collaborative episode, Olivia highlights NotebookLM's persistent ranking as a mild narrative violation against expectations that interest had faded.
Hardest push from the host ▶ 13:11 Probing for mobile vibe-coding entriesJustine explicitly pushes on whether vibe coding has made any inroads into mobile, prompting Olivia to clarify that no mobile vibe coding apps have reached top rankings yet.
Biggest teaching moment ▶ 11:18 Vibe coding 100%+ net revenue retention dataOlivia educates the host by introducing proprietary Consumer Edge cohort data showing 100%+ revenue retention in the first three months for leading vibe coding platforms.
The host holds their own ▶ 8:17 Mapping Chinese AI distribution mechanismsJustine demonstrates deep industry expertise by mapping out how Chinese AI models bypass direct distribution by leveraging U.S. consumer aggregators and developer API platforms.
the scores for every segment, with the reasoning behind each
| Chapter | Topic | The host as informed peer | Guest teaching | Guest disagreement | The host pushing back | Why |
|---|---|---|---|---|---|---|
| New Entrants, Mobile Volatility, and List Stabilization | 3 | 4 | 0 | 0 | Justine frames the discussion around category shifts away from traditional chat and creative tools toward vibe coding. Olivia provides specific data regarding list turnover, noting 11 new web entrants and the impact of iOS app store policy changes on ChatGPT clones. | |
| Google's Expansion Across Independent AI Web Properties | 3 | 4 | 0 | 0 | Olivia details how Google's domain restructuring allowed four unique properties to be ranked, including Gemini and NotebookLM. Justine adds context regarding specific usage of Google Labs tools like VO3 and Whisk. | |
| Chinese AI Applications and International Export Strategies | 6 | 4 | 0 | 0 | Justine demonstrates strong domain knowledge, explaining how Chinese video models like Seedance outscore VO3 and describing U.S. distribution layers through aggregators like Crea/Hydra and Fal/Replicate. Olivia breaks down domestic Chinese usage versus export products. | |
| Vibe Coding Economics, Revenue Retention, and Traffic Dynamics | 5 | 4 | 0 | 0 | Olivia shares Consumer Edge data demonstrating 100%+ net revenue retention for vibe coding tools like Lovable. Justine contextualizes this data, noting that high retention points to underlying enterprise prototyping usage and custom domains masking traffic. | |
| The Consumer AI All-Stars: Five-Time List Entrants | 6 | 3 | 0 | 0 | Olivia reveals the 14 five-time list entrants, noting over half rely on third-party models. Justine extends the analysis by detailing non-data network effects and explaining the bottom-up prosumer-to-enterprise adoption sales cycle. | |
| Historical Evolution Across Five Report Editions | 4 | 3 | 0 | 0 | Olivia reflects on the transition from early list volatility to current category normalization. Justine adds historical color about angry founder calls resulting from data provider reporting lags in early report cycles. | |
| Future Predictions and Emerging Consumer AI Categories | 5 | 3 | 0 | 0 | Olivia highlights Grok's debut at #4 on web and predicts prosumer productivity expansion. Justine connects foundation model leaps in math and logic directly to resolving hallucination bottlenecks in app-layer productivity tools. |
Statements from this episode (21)
Total usage reflects AI consumer interest better than revenue
“Captures usage, not revenue, so it's not paid usage only, it's all usage, even people who are using the products for free, which to me gives an even better, and I think kind of more real sense of what is capturing consumer imagination in AI.”
New entrants to top AI web apps list drop to 11
“But on the web list, we still did have 11 new names. Sounds like a lot out of 50, but it's actually a big reduction from, I think it was 17 new names six months ago. Which to me reflects the fact that the ecosystem is starting to stabilize a little bit more.”
Three new AI companionship apps enter a16z's Top 100 list
“Three new companionship names. So, Juicy Chat, Joy, and Our Dream all made the list, alongside a bunch of others that have made the list before. So, Character was the biggest name, Janitor, Spicy Chat, Polly Buzz, Crush On, Adot, and Candy AI were all on this …”
Google Gemini reaches half of ChatGPT's mobile app traffic
“Gemini was number two. It was right behind ChatGPT. It has just about 10% of ChatGPT's traffic on web, but it's much closer on mobile. It's like half of ChatGPT's traffic on the mobile app, which is interesting. It's mostly Android users, which I think you wou…”
Google AI Studio ranks in top 10 consumer AI web apps
“The other big debut for Google was a little more surprising to me. It's actually AI Studio, which is Google's developer-facing sandbox. So that's where engineers go when they want to build things and test out Google models. So that hit in the top 10 as well.”
Google NotebookLM sustains flat or growing traffic until late summer
“And then right below the top 10 was Notebook LM. At number 13. And I take your wow as a little bit of a surprise, which was a surprise to me too, and I think a bit of a narrative violation that they went viral almost a year ago. In September of 2024. But they'…”
Google Labs traffic spikes 15% upon Veo 3 release
“But my hunch is it's largely VO three, because traffic to that To the Google Labs spiked 15% in the month VO three was released, which is really exciting.”
Alibaba, ByteDance, and Moonshot AI assistants rank in top 20
“Quark, which is Alibaba's AI assistant on both web and mobile. Daobao, which is ByteDance's AI assistant on also on both web and mobile. And then Kimi, which is another general AI assistant from Moonshot AI. Each of those ranks, actually all of them in the top…”
ByteDance's Seedance video model outscores Google's Veo 3 on benchmarks
“ByteDance has also been doing a lot more around things like Seedance which is a really great video model they released that actually outscores VO-III on some of the arenas.”
Manus hits $90M ARR with top traffic from Brazil and US
“Manus is a great example. They just announced a ninety million annualized run rate. And if you look at their traffic, it's number one Brazil, number two US, and number three China.”
Lovable announces $100 million in annual recurring revenue
“Lovable recently announced a hundred million dollars in ARR.”
Vibe coding platforms see over 100% revenue retention in first months
“And we put some of this in the report, but what we found for many of the leading vibe coding platforms, Is they actually see a hundred percent or above, ah, revenue retention in the first three months, and then maybe flatten out, ah, a bit below, ah, a hundred…”
Vibe coding creation tools see higher traffic than produced applications
“And in both cases, the people making things traffic is much higher than the traffic to the things that they are making.”
Most top consumer AI apps rely on third-party models
“More than half of them are actually hosting or using other people's models or are model aggregators.”
Hugging Face and Civitai show non-data AI network effects
“Some of them are also starting to see what I would say are sort of the first kind of non-data network effects in AI. Especially things like Hugging Face and Civit AI, where, you know, you have these communities where people are hosting tons of models or lauras…”
Shared workflow assets create defensible enterprise AI moats
“These consumer AI properties are almost graduating to enterprise usage with things like team plans, templating, you know, design libraries, and then it becomes something that can spread organically through a team and that a whole organization can benefit from.…”
Over half of top consumer AI apps turned over in early reports
“The first list or two was total chaos in the sense of half the list was new every time, if not more.”
Only three AI app categories had sustained usage two years ago
“Companionship and general LLM assistance, like ChatGPT, and then creative tools were kind of the three categories that were working two years ago. And anything else that popped up on the list, in many cases, was pretty random, and would spike one month and be …”
xAI's Grok debuts at number four on top AI web apps list
“And Grok debuted, actually, in the top five at number four on the web list.”