In the late 1990s, Microsoft’s Internet Explorer crushed Netscape Navigator by bundling the browser with Windows and giving it away for free. Two decades later, Google Chrome pulled ahead with speed, extensibility, and better adherence to open web standards. Today, Chrome holds over two-thirds of the browser market. But a new front is opening up, around the potential impact of artificial intelligence.
Perplexity CEO Aravind Srinivas made the stakes clear on X: “I had reached out to Chrome a long time ago to offer Perplexity as a default search engine option. They refused. [So], we decided to build the Perplexity Comet browser.” The new AI arms race in browsers is about control over user data, something that underpins search, advertising, and increasingly, AI personalization.
A Quick Breakdown of the Emerging AI Browsers
- Comet (Perplexity): Chromium-based with full assistant, agentic actions, and strong user feedback. Emphasizes task automation and privacy.
- OpenAI (upcoming): Will include native ChatGPT and Operator agents. To be built on Chromium, aimed at ChatGPT’s existing user base.
- Brave with Leo AI: Offers both cloud and local AI via Ollama. Strong on privacy, but narrower in scope than Comet.
- Strawberry: Uses autonomous AI companions for browser-based task execution. Focused on productivity and automation.
- Dia: Offers AI chat search and programmable skills. Reception has been mixed, with performance and feature gaps cited by users.
- Opera Neon: Continues active development. Incorporating AI into its multi-platform suite, but has not yet released a browser at the level of Comet.
Most browsers with AI functionality launched since 2024 are based on Chromium. This has allowed them to maintain compatibility with Chrome extensions and to reduce switching friction for users accustomed to Google’s interface. However, simply layering a chatbot onto Chromium will not be sufficient to attract or retain users. And that is something that rivals like OpenAI are looking to move beyond.
The Incumbent- Google Chrome:
Chrome’s differentiation remains its massive distribution advantage and the ability to integrate Gemini and AI features into the browser without requiring users to switch.
Google has introduced “AI Mode” in Chrome Beta for Android, with broader rollout expected. This feature brings generative AI into the new tab search bar and adds contextual assistance. As the default browser for billions of users and the primary conduit for search traffic, Chrome can implement AI features across its user base without disrupting familiarity. Its integration is incremental but increasingly central to Google’s defensive strategy.
Comet (Perplexity):
Differentiator: Its deep integration with Perplexity’s agentic AI to automate complex, multi-step workflows directly within the browser.
Built on Chromium, Comet features an AI assistant capable of summarizing content, answering context-aware questions, and executing tasks such as form-filling, booking reservations, and extracting structured data. Users have reported using it for actions like testing websites, tracking investments, and completing browser sessions autonomously. The assistant can read content across open tabs and act within websites, giving users an interface that moves beyond navigation to task delegation.
OpenAI Browser (Upcoming):
Differentiator: Its planned integration with ChatGPT and Operator agents, embedded within an interface designed from the ground up around AI-first workflows.
Also Chromium-based, the browser will feature native chat browsing, persistent memory, and action-oriented capabilities. It will run on OpenAI’s proprietary models and benefit from the company’s existing 500 million weekly ChatGPT users. Unlike third-party plugins, the standalone browser gives OpenAI direct control over user interaction, data flow, and interface design. It is expected to keep more user actions within a chat environment and reduce reliance on traditional web navigation.
Dia (The Browser Company):
Dia looks to differentiates itself through its programmable “Skills” and chatbot-style search interface, although its execution has been uneven.
Designed as a successor to the now-shelved Arc, Dia offers contextual browsing features such as tab memory and integrated chat-based controls. It supports user-defined skills that automate browser behavior. However, reviews have pointed to limitations in performance, battery usage, and platform support. With no Windows version currently available and a reduced feature set compared to Arc, Dia has struggled to build momentum despite some promising ideas.
Brave with Leo AI:
Brave’s Leo AI stands out for its privacy-first approach and ability to run large language models locally, without cloud-based processing.
Using Ollama and llama.cpp, Leo offers AI-assisted features such as summarizing content, answering questions, and engaging in contextual interactions, all while storing data locally. This model avoids sending user input to cloud servers and appeals to users who prioritize control over their data. While Leo’s scope is narrower than competitors like Comet or OpenAI’s upcoming browser, its hybrid cloud-local model and integration with RTX hardware for faster inference make it distinct.
Strawberry:
Relatively obscure, Strawberry highlights its use of AI “companions” – autonomous agents that perform multi-step web tasks based on conversational prompts.
The browser enables users to assign structured tasks like collecting contact information, writing posts, or scraping web content, all executed by its companion agents. These agents can open and navigate multiple tabs, return results in formatted tables, and operate with minimal user input. Designed with a focus on productivity and automation, Strawberry positions itself as a tool for power users looking to offload repetitive browser-based tasks.
Opera Neon:
Opera continues to differentiate itself through incremental AI features added to its long-standing multi-platform browser.
Unlike some of the newcomers building AI-first products, Opera is integrating AI into an existing framework, experimenting with tools such as contextual summarization, chat features, and sidebar assistants. While not yet at the level of agentic browsers like Comet or Strawberry, Neon remains in active development, aiming to remain relevant through a mix of UI innovation and gradual AI feature rollouts. Its continuity and brand familiarity give it a different posture in the evolving browser market.
What Determines a Winner: Our Take
Historically, browser adoption was driven by distribution. Internet Explorer succeeded by being bundled with Windows. Chrome succeeded with speed, frequent updates, and web standard compliance. These shifts required users to believe the new option was significantly better than what they had, and often relied on ease of access and default settings.
Today, the barriers to switching are higher. Chrome users benefit from a tightly integrated Google ecosystem: syncing across devices, storing passwords, bookmarks, history, and search habits. Most people don’t change browsers unless the new one offers clear, transformative utility.
AI may qualify. But only if it establishes itself as a necessity, and that’s largely out of the hands of all but those with the most backing. Comet aims to do that by turning browsing into task execution. OpenAI’s browser intends to collapse search, navigation, and action into a single interface. These are departures from the traditional tab-centric model. The value is in whether they really can reduce time spent navigating or increase what users can accomplish without switching context.
A structural advantage remains with Google. Chrome is still the default on most Android devices and desktops, and many of the AI features others are touting can be replicated or absorbed into Chrome itself. The difference comes down to execution, and to whether users care more about capability, privacy, or compatibility.
Unlike earlier browser wars, the current wave doesn’t center on rendering engines or standards compliance. Most AI browsers use Google’s Chromium base. This consolidates power in Google’s hands even as rivals attempt to build on its infrastructure. Chrome’s market share continues to benefit from this consolidation, regardless of how advanced its competitors become.
That’s part of the calculus behind OpenAI’s decision to build its own browser rather than integrate ChatGPT as a plugin. According to reports, OpenAI wanted full control over the data layer, product design, and AI integration, none of which are possible when relying on another company’s framework for distribution or permissions. The company has also expressed interest in acquiring Chrome, should regulatory efforts force its divestiture.
For now, Chrome is not for sale. But the DOJ’s antitrust scrutiny has opened the door for competitors. If users begin shifting toward AI as their primary interface to the web, and if new browsers can prove material gains in time saved or productivity, Chrome’s grip may begin to loosen.
