Quelle:
https://www.ft.com/content/7ffea5b4-e8bc-47cd-adb4-257f84c8028b?syn-25a6b1a6=1
https://ramp.com/velocity/ai-index-march-2026 Ramp Economics Lab (ramp.com/data) (unter Verwendung von Daten von Rampenrate (ramp.com/vendors); Karten- und Rechnungszahlungsdaten von über 50.000 Unternehmen auf der Ausgabenplattform von Ramp. Gleitender 7-Tage-Durchschnitt. Ausgenommen sind Ausgaben für KI-Modelle, die nicht von OpenAI und Anthropic stammen.)
Von Mundane-Wrongdoer275
9 Kommentare
Source:
[https://www.ft.com/content/7ffea5b4-e8bc-47cd-adb4-257f84c8028b?syn-25a6b1a6=1](https://www.ft.com/content/7ffea5b4-e8bc-47cd-adb4-257f84c8028b?syn-25a6b1a6=1)
[https://ramp.com/velocity/ai-index-march-2026](https://ramp.com/velocity/ai-index-march-2026) [Ramp Economics Lab (ramp.com/data)](https://ramp.com/data) (using data from [Ramp Rate (ramp.com/vendors)](https://ramp.com/vendors); card and bill pay data from 50K+ businesses on Ramp’s spend platform. 7-day moving average. Excludes spend on AI models not from OpenAI and Anthropic.)
Claude Code launching and Anthropic adding a billion in annualised revenue every week this year explains a lot of this movement.
Unsurprising, Opus 4.6 is miles ahead of GPT-5.4
OpenAI signing deals with the Pentagon and pushing into government contracts while simultaneously trying to sell to enterprise businesses is creating a trust problem with a lot of corporate buyers
Sam Altman issuing a code red and telling employees to refocus while this chart is playing out in the background is a specific kind of pressure that does not get easier to manage
claude doesn’t write a 5 paragraph essay with bolded, emoji-containing headings when you ask a yes or no question, so yeah this tracks
Both companies are burning billions and neither is profitable, the race to capture business customers is essentially a race to see who runs out of runway first
It’s not just the quality of the LLM itself, which is better with anthropic – I feel like all the tools they develop are better, and they deploy a lot more than openAI
I mean yeah. PR aside, Claude just objectively works better. Much more reliable and understands prompts more clearly.