Research
Claude vs ChatGPT: The Rivalry That Is Reshaping AI Search, April 2026
Anthropic hit $30 billion in annualized revenue while OpenAI scrambled to match Claude Code with a new Pro tier. The rivalry between these two platforms is now directly shaping how brands get discovered in AI search.
The last two weeks delivered one of the sharpest competitive escalations in AI history. Anthropic passed OpenAI in revenue, released a model that finds zero-day vulnerabilities across every major operating system, and watched Claude become the number one app on the App Store. OpenAI responded with a $100-per-month Pro tier aimed directly at Claude Code’s dominance in enterprise coding.
For anyone tracking AI visibility, these moves are not just corporate drama. They determine which AI systems mediate the most content consumption, which platforms developers build on, and which search surfaces your brand needs to appear in.
Here is everything that happened and why it matters.
Anthropic passes OpenAI in revenue at $30 billion ARR
Anthropic hit $30 billion in annualized recurring revenue in April 2026, officially passing OpenAI’s $25 billion ARR. The gap is particularly striking given that Anthropic reportedly spends four times less on model training than OpenAI.
The revenue composition tells the real story. Anthropic generates 80 percent of its revenue from business customers, compared to OpenAI’s more consumer-heavy mix. Claude Code alone is now a multi-billion-dollar product line, with run-rate revenue exceeding $2.5 billion as of February, a figure that more than doubled since the start of 2026.
Anthropic now owns 54 percent of the enterprise coding market, making it the default AI coding tool for most large organizations.
Why it matters for visibility: The platform that dominates enterprise adoption shapes how professional content is consumed. With 80 percent of Anthropic’s revenue coming from business users, Claude is increasingly the AI system through which enterprise decisions are informed. If your content targets B2B audiences, Claude visibility is no longer optional.
Claude Mythos Preview and Project Glasswing
Anthropic announced Project Glasswing, a cybersecurity initiative built around Claude Mythos Preview, a frontier model so capable at finding software vulnerabilities that Anthropic chose not to release it publicly.
Over the past few weeks, Claude Mythos Preview identified thousands of zero-day vulnerabilities across every major operating system, every major web browser, and a range of other critical software. Anthropic engineers with no formal security training asked Mythos Preview to find remote code execution vulnerabilities overnight and woke up the next morning to complete, working exploits.
Anthropic committed $100 million in model usage credits to the initiative and granted early access to Apple, Microsoft, Google, AWS, CrowdStrike, NVIDIA, JPMorgan Chase, Cisco, and the Linux Foundation. Claude Mythos Preview scored 97.6 percent on capability benchmarks, the highest of any model released in 2026.
Why it matters for visibility: Mythos is not a search product, but it signals the capability gap between frontier models and everything else. When a model can autonomously find and exploit software vulnerabilities that skilled humans miss, the same reasoning capability applied to content retrieval and citation means AI systems will become even more selective about which sources they trust and cite.
OpenAI launches ChatGPT Pro at $100 per month
OpenAI launched ChatGPT Pro, a new $100-per-month subscription tier that offers five times more Codex usage than the $20-per-month Plus plan. Through May 31, Pro subscribers get up to 10x Plus usage on Codex.
The move is a direct response to Claude Code’s market dominance. OpenAI’s Codex, powered by the GPT-5.3-Codex model, is positioned as the counter to Anthropic’s coding agent. The pricing structure, $20 for Plus, $100 for Pro, mirrors the tiered approach Anthropic uses with Claude Pro and Claude Max.
OpenAI also expanded ChatGPT Business and Enterprise with Codex-only seats on pay-as-you-go pricing, no rate limits, and clearer token-based billing.
Why it matters for visibility: The coding tool war is a proxy for platform adoption. Developers build products, integrations, and content systems on the platform they code with. If Claude Code owns the enterprise coding market, those developers build AI features that query Claude, not ChatGPT, for information. Platform adoption among developers determines which AI system becomes the default information retrieval layer.
Claude becomes the number one app on the App Store
Claude became the most downloaded free app on Apple’s U.S. App Store, surpassing ChatGPT. Claude’s user base has grown over 60 percent since January 2026.
The surge was fueled in part by public attention around Anthropic’s Pentagon standoff. Anthropic’s refusal to allow Claude to be used for autonomous weapons or mass surveillance, and the Pentagon’s retaliatory blacklisting, generated significant media coverage and public support.
Why it matters for visibility: Consumer adoption shifts AI search market share. More users on Claude means more queries flowing through Claude’s search and citation system. Content that is visible in Claude’s answers reaches a growing audience that was previously only accessible through ChatGPT or Google.
Claude Code auto mode and Managed Agents go live
Anthropic shipped two significant platform capabilities this month. Claude Code auto mode lets the AI decide which actions are safe to take autonomously, currently in research preview for Sonnet 4.6 and Opus 4.6.
Claude Managed Agents entered public beta, providing a fully managed agent harness for running Claude as an autonomous agent with secure sandboxing, built-in tools, and server-sent event streaming.
Separately, Anthropic launched Claude Computer Use and Dispatch. Users can message Claude a task from their phone, and the agent opens apps, navigates browsers, and completes work on a Mac. Available to Claude Pro and Max subscribers.
Why it matters for visibility: Agentic AI changes how content is consumed. When Claude agents browse, read, and act on web content autonomously, your pages need to work for machine consumption, not just human reading. Structured data, clean information architecture, and machine-readable formats become critical for being discovered and used by AI agents.
ChatGPT upgrades shopping and commerce features
ChatGPT improved its shopping experience with richer product results, conversational browsing, image-based similar item search, and side-by-side comparisons with price, reviews, and features.
This builds on OpenAI’s Agentic Commerce Protocol expansion, with Target, Sephora, Nordstrom, Best Buy, and The Home Depot now integrated for product discovery and checkout directly inside ChatGPT.
ChatGPT also added mobile sidebar improvements and optional device location sharing for more localized recommendations, news, and weather.
Why it matters for visibility: ChatGPT is becoming a shopping search engine. Product brands that are not optimized for ChatGPT’s discovery layer, through structured product data, reviews, and ACP integration, are losing a growing commerce channel. The image-based search feature means product photography quality now directly affects AI discoverability.
Claude outages hit back-to-back on April 7-8
Claude experienced back-to-back service disruptions on April 7 and April 8, with hundreds of users reporting login failures, chat errors, and degraded performance. The April 8 outage came less than 24 hours after the previous incident was resolved.
For a platform that just passed $30 billion ARR and handles enterprise workloads at scale, reliability is a competitive vulnerability. OpenAI has maintained stronger uptime through 2026, and service disruptions give enterprise buyers a reason to hesitate on deeper Claude commitments.
Why it matters for visibility: Service reliability affects which AI platform enterprise teams default to for daily workflows. If Claude is down, those queries flow to ChatGPT or Google, temporarily shifting which AI system is consuming and citing content. For brands tracking AI visibility, platform-specific monitoring needs to account for outage-driven volatility.
Pentagon blacklist stays in force
The D.C. Circuit denied Anthropic’s emergency stay on April 8, allowing the Pentagon’s designation of Anthropic as a supply chain risk to remain in effect. Expedited oral arguments are set for May 19, 2026.
Anthropic argues the blacklisting is retaliation for its refusal to allow Claude for autonomous weapons or mass surveillance. The Pentagon’s position effectively bars federal agencies from using Claude, pushing government AI workloads toward OpenAI and Google.
Why it matters for visibility: Government and defense-adjacent organizations cannot use Claude, which means content serving those audiences needs to be optimized for ChatGPT and Google AI surfaces. The outcome of the May hearing could reshape the entire government AI procurement landscape and determine which platforms mediate information access in the public sector.
Bottom line
Three patterns emerge from the last two weeks:
-
The coding tool war determines AI platform adoption. Claude Code owns 54 percent of enterprise coding. OpenAI launched a $100 Pro tier to compete. Whichever platform wins developers wins the default AI surface for content discovery. Your content needs to be visible across both.
-
Agentic capabilities are expanding on both sides. Claude Managed Agents, Computer Use, and auto mode on one side. ChatGPT shopping, Codex, and commerce integration on the other. Both platforms are moving from chat interfaces to autonomous task execution, which means more content consumption without human browsing.
-
Cross-platform AI visibility is now essential. Claude’s 60 percent user growth and number one App Store position make it impossible to ignore alongside ChatGPT. Brands measuring AI visibility on only one platform are seeing an incomplete picture. The rivalry guarantees that neither platform will dominate exclusively, which means your content needs to perform in both.