Claude Code was the most-used coding agent among businesses on Ramp in October 2026, according to Ramp Rate pages last updated on 12 October 2026: 87% of businesses that bought a coding agent in the past 12 months bought Claude Code. Codex was second at 40% and Cursor third at 35%.
The same Ramp Rate data names Anthropic the most-used foundational LLM vendor at 80%, just ahead of OpenAI at 78%, and lists the leader in 14 AI categories, from Perplexity at 95% in search and research to Fireworks AI at 27% in model serving.
Business Pill · SWITCHING COSTS
A one-minute explainer of switching costs: what it costs a customer to move from one supplier to another. It teaches the general idea only and says nothing about any company in this story.
The key insight: As we read it, Ramp’s data counts who pays, not who tries: by that measure Claude Code is bought by 87% of businesses buying a coding agent, while in foundational LLMs Anthropic at 80% and OpenAI at 78% are each bought by most category buyers, so many businesses pay for both.
Ramp Rate measures what businesses actually pay for. Its methodology says the data comes from aggregated, anonymized corporate card and bill pay transactions covering billions of dollars in spend across more than 70,000 American businesses. Adoption rate is the share of businesses that purchased from a vendor, out of all businesses that made purchases in that software category over the past 12 months.
That makes these shares of buyers, not shares of dollars. A business that pays for Claude Code and Codex counts for both, which is why the coding-agent shares add up to more than 100% by our arithmetic.
In coding agents, Ramp says Claude Code also attracted the most companies switching from alternatives, while Codex is the fastest-growing vendor, averaging +1.6 percentage points per month. The share of businesses using any coding agent was 30.3% in October 2026, down 1.0 percentage points from the previous month, according to the page.
The most-used vendor in each of 14 AI categories on Ramp Rate and its adoption rate: the share of businesses that bought from it, out of all businesses that bought in that category over the past 12 months. Ramp Rate category pages, last updated 12 October 2026, covering more than 70,000 American businesses on Ramp.
Anthropic and OpenAI, Almost Level
In foundational LLMs, Ramp lists Anthropic at 80% and OpenAI at 78%, followed by xAI at 10% and Mistral AI at 1%. Ramp says Anthropic is the fastest-growing vendor in the category, averaging +0.9 percentage points per month, and attracted the most companies switching from alternatives.
The share of businesses that rely on foundational LLM tools reached 65.2% in October 2026, up 1.1 percentage points from the previous month, according to the category page. Our earlier coverage of Ramp’s token-spend data tracked the same two labs from the spend side.
The 14 Category Leaders
Ramp Rate names a most-used vendor in each AI category. Search and research AI: Perplexity AI at 95%. AI dictation: Wispr Flow at 94%. Coding agents: Claude Code at 87%. Foundational LLMs: Anthropic at 80%. Voice AI models and infrastructure: ElevenLabs at 77%. AI meeting notetakers: Granola at 64%. AEO: Profound at 53%.
AI support bots: Intercom at 46%. Vector databases: Pinecone at 45%. Neocloud: Runpod at 44%, ahead of Modal at 40%; Runpod’s revenue run rate was in our earlier piece. Image and video AI: Midjourney at 41%. Wholesale GPU capacity: Andromeda Cluster at 36%. AI-first GTM: Attention at 34%. Model serving and inference: Fireworks AI at 27%.
The spread is wide. In search and research and in dictation, the leader is bought by more than nine in ten category buyers, with the next vendor at 9% in both. In model serving, the leader reaches just over a quarter.
The Structural Read
As we read it, the coding-agent category is led by a lab’s own tool: Ramp lists Claude Code at 87% and Codex at 40%, ahead of Cursor at 35%, and says Claude Code drew the most companies switching from alternatives.
As we read it, concentration differs by category: the leaders in search and research and in dictation are bought by more than nine in ten category buyers, against just over a quarter for the model-serving leader, according to Ramp Rate.
As we read it, these are shares of buyers on one platform, so they say who businesses pay rather than how much; the coding-agent shares sum to more than 100% by our arithmetic because businesses can buy several tools.
Ramp Rate, Coding Agents category page, last updated 12 October 2026
“Claude Code was the most-used vendor in the Coding Agents space as of October 2026, making it the top choice among businesses using tools in this category.”
Three Implications
CODING-TOOL VENDORS Ramp says Codex is the fastest-growing coding agent, averaging +1.6 percentage points per month, while Claude Code attracted the most companies switching from alternatives.
MODEL LABS Ramp lists Anthropic at 80% and OpenAI at 78% in foundational LLMs and says Anthropic is the fastest-growing vendor in the category, averaging +0.9 percentage points per month.
AI INFRASTRUCTURE BUYERS In the infrastructure categories Ramp lists Runpod at 44% in neoclouds, Andromeda Cluster at 36% in wholesale GPU capacity and Fireworks AI at 27% in model serving.
The framework puts it this way: “What emerges is a barbelled market structure where software companies must choose between racing to the commoditization floor or climbing to the embedding ceiling.”
As we read it, Ramp’s category pages show both ends: categories where one vendor is bought by more than nine in ten category buyers, and others where the leader reaches about a quarter.
The top four coding agents by adoption rate on Ramp Rate’s Coding Agents page, last updated 12 October 2026: Claude Code 87%, Codex 40%, Cursor 35%, CodeRabbit 5%. Businesses can buy more than one tool, so the shares sum to more than 100% (our arithmetic).
What Is Not Established
Ramp’s figures cover businesses that use Ramp, mostly American companies, and count buyers rather than spending. They do not show how much each business spends, how many seats it buys or how heavily it uses a tool. Free tiers and purchases made outside Ramp cards and bill pay do not appear. We read the 14 Ramp Rate category pages directly; we did not find a separate Ramp post publishing the graphic.
On Ramp’s data, Claude Code is bought by 87% of businesses buying a coding agent and Anthropic by 80% of those buying a foundational LLM, with OpenAI close behind at 78%. The leaders range from Perplexity at 95% to Fireworks AI at 27% across 14 AI categories.
A note on sourcing. On 14 October 2026 we read the 14 Ramp Rate AI category pages, each last updated 12 October 2026, and their methodology on ramp.com. Every figure here was checked against those pages. Shares are of buyers on Ramp, not of spending. Nothing here is investment advice.
Gennaro Cuofano is a CRO and tech executive who has worked in AI since late 2015, starting with bringing NLP, natural-language generation, voice and chatbot products to the marketing industry. His background is in law and finance: he holds a Master's degree in Law and an International MBA with an emphasis on corporate finance (LUISS Business School and the University of San Diego, 2012), and worked as a financial analyst at a real-estate investment firm in San Diego and as an assistant controller. His work focuses on business model strategy, business engineering and, more broadly, structural analysis: how companies make money, read from their own filings. He created FourWeekMBA and leads research there and at The Business Engineer, his newsletter on AI and business strategy, with over 95,000 subscribers and more than 1,000 published analyses. His writing has also appeared on Entrepreneur, HackerNoon and Search Engine People. Find him on LinkedIn and Substack. See how we source.
Scroll to Top
Discover more from FourWeekMBA
Subscribe now to keep reading and get access to the full archive.
