Part of: this research — view all 4 rows
Scope and Starting Position There is no clean, published revenue-share series for AI coding assistants, so this estimate requires rebuilding the market from estimated late-2026 run-rates. The denominator comprises Claude Code, OpenAI’s coding products (Codex/ChatGPT Work), Cursor, GitHub Copilot, and Google’s coding surfaces. Across these five, annualized revenue is roughly $40 billion today. Claude Code sits at an estimated $15–19 billion (derived from a roughly 22% share of Anthropic's $81 billion total run-rate), placing its current share in the 42–48% range. OpenAI's coding products follow at roughly $9–16 billion, supplemented by Cursor at $5–6 billion (annualizing the $608 million SpaceX reported from mid-August to September 30 Cursor Statistics 2026: Revenue, Users and Pricing), GitHub Copilot at $1–3 billion, and Google under $2 billion.
Usage Proxies vs. Revenue Reality Proxy metrics support Anthropic's current leadership but require adjustment to reflect revenue. The August 2026 JetBrains survey found Claude Code at 39% penetration, versus Copilot at 21%, Codex 16%, and Cursor 12%. Menlo Ventures’ December 2025 survey assigned Anthropic 54% of enterprise coding spend, though as an Anthropic investor, their figures warrant a slight discount. Crucially, usage share and revenue share diverge sharply. Claude Code likely commands higher average revenue per user (ARPU) due to heavy, usage-billed agentic spend. Conversely, Codex and Copilot are heavily bundled into broader subscriptions (e.g., ChatGPT, Microsoft enterprise deals), making exact attribution difficult but ensuring massive distribution.
Trends for 2027 Anthropic enters 2027 with a formidable capability advantage to defend its premium share. Claude Fable 5.2 leads the Artificial Analysis Intelligence Index with a score of 64.6, ahead of Gemini 4 Carbon (60.0) and GPT-6 Astra (56.9). However, intense competitive pressure will likely erode Anthropic's share modestly over the year. Codex is growing rapidly month-on-month Claude Code Reaches $15.1B in Tracked ARR as Codex ... as OpenAI pushes enterprise sales, GitHub Copilot is accelerating following its shift to usage-based billing, and Cursor is expanding explosively backed by SpaceX capital and its own in-house models. Google's migration of Code Assist customers to Antigravity also maintains competitive heat. Weighing Claude Code revenue trajectories against these pressures, agentic workload growth should help sustain Anthropic's revenue even as aggressive competitor bundling limits its share. I expect Anthropic's share to drift slightly downward from its current position, settling around a median of 44.5% for the full calendar year.
Accounting and Resolution Uncertainties A wide 80% interval (31% to 60.5%) is necessary due to severe accounting ambiguities, though the balanced risks between agentic workload growth and aggressive bundling favor somewhat tighter tails. The middle 50% of my distribution spans 38% to 52%. Anthropic recognizes cloud-marketplace sales gross (which make up 44% of its revenue), whereas OpenAI books partner sales net. Mixing these unadjusted corporate run-rates inherently overstates Anthropic’s relative position. Furthermore, it is unclear exactly how a resolver will classify coding done via Anthropic's general API or through Claude subscriptions versus the standalone Claude Code product. If the resolution leans on strict, like-for-like net revenue and Cursor captures massive enterprise share, Anthropic could fall toward 31%. Conversely, if Anthropic's capability lead translates into an overwhelming enterprise ARPU advantage, or if the resolver counts all Anthropic coding-related API spend (as Menlo does), share could push past 60%.
Weighing this against Claude Code revenue trajectories , the distribution was slightly tightened to reflect balanced risks between agentic workload growth and aggressive bundling by competitors.
Ask a followup
Sign in to run · $20 free credit, no card · every claim cited