A look at Anthropic's success in corporate AI, which makes up ~80% of its revenue; a July report put its enterprise API market share at 32%, above OpenAI's 25%
and it's all thanks to this one problem OpenAI can't solve Maria Garcia / Implicator.ai : Anthropic's finance push lands while revenue closes in on OpenAI Samreen Ahmad / Tech in Asia : Will India's trade tensions spill over to OpenAI, Anthropic? Webb Wright / ZDNET : Anthropic expands Claude for Financial Services with a beta Claude for Excel integration, additional data connectors, and new pre-built Agent Skills X: @peterwildeford : OpenAI is mopping up on free users but losing out on enterprise... this seems not the direction you want to be going in as an AI company John Baris / @dogancanbaris : @StefanFSchubert I see no way Openai can compete with them because Google backs Anthropic. OpenAI is stuck with consumers who won't ever pay money when Google has free search engine and ai mode @doctoryev : @StefanFSchubert slope difference between Anthropic and Google will get super interesting. @chachangfan : @StefanFSchubert the most shocking thing about this is meta having anywhere close to 10% market share @thenohtus : @StefanFSchubert I thought only programmers knew about claude, guess not Roy Derks / @gethackteam : @StefanFSchubert Google doubling their market share surprises me more @_valsai : @StefanFSchubert Enterprise buyers prioritize capability, and Anthropic has demonstrated exceptional ability in the domains measured by our Vals Index. h/t @asafitch [image] Timbo / @timdodaro2 : @StefanFSchubert I would back up the truck if Anthropic had an ipo Garrison Lovely / @garrisonlovely : @StefanFSchubert FYI that chart is from July and is from a major investor in anthropic. Itamar Golan / @itakgol : @StefanFSchubert Worth mentioning the graph's source - Menlo Ventures - and that they're investors in Anthropic. Nathan Benaich / @nathanbenaich : @StefanFSchubert slightly biased source in that i comes from a major shareholder? Adrian / @notadrianmadame : @StefanFSchubert is the anthropic vertex stake counted toward anthropic or google? Ali Sherief / @zenul_abidin : @StefanFSchubert This is what happens when “do it right” beats “ship it fast.” Claude's consistency speaks volumes @muh_saad0 : the thing about anthropic going from 12% to 32% while openai dropped from 50% to 25% isn't just the crossover. it's how fast it happened. two years. that's not gradual market shift, that's enterprises fundamentally rethinking what they need. and claude dominating code generation—42% versus openai's 21%—tells you something real about production use... Neil / @itsonolo : @StefanFSchubert we know the public, commercial models are pretty close in performance between the two. so enterprise partnerships could come down to several other factors: - private / custom models - uptime SLAs - better pricing / contract terms Petri Kuittinen / @kuittinenpetri : @StefanFSchubert I find hard to believe this chart fully. It implies that just 4 companies (OpenAI, Anthropic, Google and Meta) have the entire 100% enterprise LLM API market to themselves. Where is XAI Grok? DeepSeek? Qwen? Kimi? ZAI GLM? Mistral? Latest OpenRouter TOP20 rankings look different [image] David Logger / @david_logger : @StefanFSchubert Interesting is there any data on industries where it's happened or just overall? Oscar Le / @oscarle_x : @StefanFSchubert I would be very surprised if Anthropic get that much market share, given that input token price of Sonnet is 3x more expensive than GPT-5 and Gemini-pro. ($3.75 vs $1.25) And enterprise are very input token heavy Gary Marcus / @garymarcus : OpenAI invincible, right? Right? AshutoshShrivastava / @ai_for_success : Anthropic is winning, and it looks like Google is also about to take over OpenAI soon in API market share. [image] Stefan Schubert / @stefanfschubert : Anthropic has overtaken OpenAI in enterprise large language model API market share [image] LinkedIn: Matt Murphy : Anthropic leading and pulling ahread in AI for the enterprise... via The Wall Street Journal “Heard on the Street: Anthropic is a less-flashy competitor to OpenAI … Bluesky: Jesse Felder / @jessefelder : 'For all of OpenAI's spending, Anthropic has shown it is as good or better in AI arenas that companies care about. Vals AI ranks the latest version of Anthropic's LLM Claude as top in a business-focused benchmark that brings together finance, legal and coding tasks.' www.wsj.com/tech/ai/anth...
Context & Ripple Effects
Anthropic’s enterprise position follows a product arc aimed at more dependable and operationally useful models: Claude 3’s effort to reduce hallucinations was followed by Claude 4 features and the broader availability of its coding agent. The company’s earlier funding round, reported alongside roughly $1.2B in annualized revenue, gave that enterprise push a larger base to build on.
The reported API-share lead matters because corporate customers account for about 80% of Anthropic’s revenue. Its financial-services expansion, including an Excel integration and data connectors, shows the competition moving beyond general-purpose chat toward deployment in established work systems.
First-order effects
- Anthropic enters enterprise API sales with a reported 32% share versus OpenAI’s 25%, strengthening its position with the corporate customers that supply most of its revenue.
- OpenAI faces a more immediate enterprise challenge: it must defend API accounts while Anthropic packages Claude around coding and finance-oriented workflows.
Second-order effects
- Enterprise buyers gain greater leverage to compare model vendors by workflow fit, particularly for coding and financial work, rather than treating a single general-purpose model provider as the default.
- Microsoft’s reported use of Anthropic models for some Office 365 features after task-specific testing underscores how enterprise software platforms can become a distribution channel for competing frontier models.
Third-order effects
- If enterprise adoption continues to split by task and integration, frontier-model competition may be decided less by broad consumer reach than by reliability, tooling, and procurement fit inside business software.
- The market is shifting toward an AI procurement phase in which large customers can sustain multiple model suppliers; whether this produces durable specialization or renewed platform consolidation remains unresolved.
The trend: Enterprise AI is evolving from a contest for general model visibility into a contest for workflow-specific performance, integrations, and durable corporate distribution.