Chatbot Wars: A New Stack
How ChatGPT won with breadth, Grok and Gemini stayed relevant through integration, and the rest got squeezed.
A short time ago, in this very galaxy...
OpenAI launched ChatGPT, and the AI arms race ignited. Billions in venture funding followed. Startups and tech giants scrambled to stake their claim in the consumer AI chatbot market: Claude, Google Bard (later renamed Gemini), Perplexity, DeepSeek, and Grok all entered the arena. But now that the dust has settled, a few contenders have emerged as frontrunners—and if my monthly subscription log is any indicator, the market is speaking clearly.
I subscribed to nearly every AI chatbot I could get my hands on. Over time, most of them fell off. Today, I personally pay for just two—ChatGPT and Grok—while at work I use Gemini through my corporate account. My personal stack mirrors what I think is happening in the broader market: ChatGPT dominates as the standalone default while Grok and Gemini remain relevant through deep integration with existing products.
Why did these products win?
Success in the consumer AI chatbot space has been about positioning rather than just model quality. And in this case, it largely came down to two vectors:
1) Broad vs. Focused Toolkit
Broad Toolkit: AI chatbots offering a comprehensive range of features—general conversation, search, image gen, coding, deep research—enabling daily-driver usage without needing to context-switch tools.
Focused Toolkit: AI chatbots that go deep in a narrow domain (e.g. writing, search, or code) but lack breadth of tools
2) Standalone App vs. Integrated Experience
Standalone App: Self-contained chat apps built to serve as the AI operating system for your digital life.
Integrated App: AI woven into the apps and ecosystems you're already using—Docs, X, Notion, etc.
Here’s how the major players map onto this landscape:
This matrix explains how the market shook out. If you’re a standalone app, you must have a broad toolkit or you get outcompeted by someone who does. ChatGPT became the daily driver by commoditizing the best features of every specialist tool, not necessarily by doing them better, but simply by adding them and making them good enough—search like Perplexity, personality like Claude, reasoning capabilities like DeepSeek.
Focused apps couldn’t keep up. Their unique advantages became checkboxes in someone else’s roadmap.
On the flip side, integrated apps like Gemini and Grok didn’t need to win on raw capability. Their advantage came from distribution. You don’t need to convince people to use your AI chatbot if it’s already inside the apps they use every day, and even more so if it’s embedded directly into the user flows they already rely on. Gemini lives in Google Workspace (which also helped them break into enterprise), while Grok is built into Twitter/X. In contrast, while their integrated versions have done quite well, I don't think their standalone apps have received nearly as much adoption.
A Callout on Claude
Claude deserves special mention. For a while, it felt like the most thoughtful AI chatbot. Great writing voice. Strong reasoning. Huge context window. But it lacked basic tools: no search, no deep research, no image generation. And that meant I couldn’t rely on it as my primary AI chatbot—I still had to turn to ChatGPT, which offered the additional tools I needed in one place. (As an example, I relied on ChatGPT's image generation while DM'ing D&D campaigns!)
As ChatGPT added more tools and memory, Claude gradually faded from my workflow. Its differentiation was eroded. And eventually, when ChatGPT launched its memory feature, I outright cancelled my Claude subscription. It might keep a loyal niche of power users, but unless Anthropic makes a bold pivot, Claude won’t reclaim ground.
How ChatGPT Maintained Its Crown
Beyond the key dynamics above, ChatGPT also executed a masterclass of defending its position in the market:
Viral Moments: ChatGPT had two viral moments, the initial launch, and then image generation where users Ghibli-fied everything. These became internet-wide events. ChatGPT landed in the middle of cultural conversations and gave people something to share, remix, and build on.
Model Quality: Good enough to stay in the top tier. While it hasn't always led on benchmarks, ChatGPT consistently delivered reliable, coherent responses across a wide range of tasks, enough to meet user expectations and build trust over time. Some of ChatGPT's features (like deep research) are also clearly ahead of the pack.
Polish and Consistency: It works, and it works the same way every time. That reliability matters. Gemini and Grok—two other major broad toolkit competitors with standalone versions—often feel unpredictable and buggy. I’ve run into authentication issues with Gemini, and my Grok subscription on Twitter doesn’t seem to sync cleanly with the standalone app. That kind of friction breaks the flow and pushes users away. I’d rather just go back to ChatGPT than troubleshoot.
Brand Strength: Thanks to its early mover advantage and viral moments, ChatGPT became the default frame for what an AI chatbot is. When people think "AI chatbot," they think ChatGPT, which creates powerful mindshare and brand gravity. In contrast, competitors like Gemini bungled their initial launch (remember the "diverse" images of US presidents?) while Grok was very late to the party.
Memory: The retention lever. With memory, ChatGPT introduced a personalized layer that made conversations more useful and relevant over time, meaningfully increasing switching costs for users who’d invested time and context.
Together, these moves cemented ChatGPT's early distribution advantage and reduced the incentive for users to switch.
What’s Next?
ChatGPT becomes the AI default. It’ll likely continue consolidating its position as the go-to AI chatbot by layering on new tools and infrastructure, such as identity (“Sign in with ChatGPT”), an ad network to scale free usage, and an agent mode. It’s shaping up to be the Google Chrome of AI—the most widely used and embedded in daily workflows.
Specialists retreat or pivot. Claude, Perplexity, DeepSeek, and others will likely shift toward narrower use cases or developer-first strategies. Expect to see more focus on API platforms, integrations with vertical tools, or enterprise offerings rather than trying to compete with ChatGPT head-on in the consumer space.
Acquisitions accelerate. Larger tech players who haven’t made as much progress on AI chatbots may start buying their way in. For example, Apple could buy Perplexity to give Siri a second life.
We’re entering phase two of the Chatbot Wars. The mass land-grab is over. Now it’s about consolidation, deeper integration, and functionality that moves beyond conversation.
Meanwhile, I’m happy to be saving $60 a month after trimming down my chatbot subscriptions!