Source:
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) (using data from Ramp Rate (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.)
Posted by Mundane-Wrongdoer275
28 Comments
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.
I have questions. The symmetry of these wiggles suggests to me that hardly anybody is **starting** to pay anybody other than these two. Is Gemini really only ever the **second** service other companies pay for? Hmmmm. I realize this chart “excludes spend on AI models not from OpenAI and Anthropic,” but that still doesn’t explain this degree of symmetry. Maybe “data from … businesses on Ramp’s spend platform” explains it because Ramp isn’t connected to anybody other than OpenAI and Anthropic? Is this really just an ad for Ramp?
Sam Altman is probably pretty angry with himself for not negotiating harder with Dario when he was leaving…
Oh who am I kidding that man is incapable of introspection.
Out of the loop, can someone give me an overview of AI companies?
I hope this isn’t bad news for openAIs’ bottom line
I wonder if we will get to know which AI was responsible for the iranian schoolgirls
So we know there are other ai vendors and we just aren’t including them.
What is ramp and why would someone use this platform to spend money when you can just go right to the vendor? So we’re missing data from folks just going to any vendor.
This says it’s only new ai spend, so any company that has ever spent money on ai (on ramp) is excluded.
It’s data but how meaningful?
Claude simply doesn’t dream things up.
It works much much better for me.
Are we watching the downfall of OpenAI right now or does it not really matter, given the government deals and the fact that everyone and their grandma stillt knows ChatGPT?
Where does gemini sit among all of this?
“Has overtaken” as in two and a half months ago.
VS Code + Claude Code
Chefs kiss
I’ve seen some interviews with Claude’s top researchers and they seem to be very aware of, actively researching, and willing to publically discuss, negative aspects of AI.
No shit, OpenAI is just bad. And keeps getting worse with every update.
OpenAI days are numbered. There is no coming back from the money furnace model they have. Most other competitors have already developed competing products, and Google, Meta and Amazon are actually profitable due to their other/core businesses (they can also put data center infrastructure for AI to other uses if need be). Sam Altman is nothing but a snake oil salesmen, whose lies about ‘AGI around the corner if only we scale it all’ are about to catch up to him. He will/should end up in jail at some point if/when the AI bubble pops and untold amounts are lost in direct investments as well as the economic shock.
Saying “we won’t let the government use our technology to create autonomous human killing machines” makes people want to support you more than the guy who doesn’t say that.
Go figure.
I mean claude.ai is more work oriented. Most people working do not require their AI to create images or videos but graphs and actual reasoning. What is scary good is it’s memory to keep track of a conversation. Feels like a genuine person. Gemini is less bullshit but forgets stuff mid convo all the time. I would rather it being less powerful but more consistent.
Everything the government touches….
https://preview.redd.it/pus14xr3l9rg1.jpeg?width=1290&format=pjpg&auto=webp&s=d4649222ff27c000c45ac000b87c8d216a104399
From same article. It seems a lot of people did not read the whole thing. It is far too early to tell a winner or where the future of AI is headed between these companies.
Only one of those companies will be getting bailed out by the Federal government