BusinessClaude

Ramp: AI spending stalls even as usage climbs

Original title: Ramp data shows AI spending stalling just as Anthropic nears its IPO

Plain language

What it is: A new report from Ramp shows that business spending on AI has flatlined, even though companies are using AI more than ever. This slowdown is caused by price cuts from AI providers and better cost management by businesses.
How it works: Businesses are keeping costs down by setting employee budgets and using cheaper AI models for simple tasks. AI vendors like OpenAI and Anthropic have also cut their prices to compete for customers.
Where it is available: This trend is happening across companies that track their expenses through the Ramp platform, particularly affecting services from OpenAI and Anthropic.
Example: Some companies are using a lightweight tool called Jev to handle simple yes or no decisions instead of using expensive models.

What it is: Ramp data indicates that enterprise AI token spend growth slowed to 1.5% over August and September, despite rising token volumes. This trend is driven by API price competition and advanced cost-optimization strategies by developers.
How it works: Organizations are optimizing costs by implementing usage budgets, setting defaults to avoid expensive models for long-running tasks, and using routers to shift workloads to cheaper light models. Vendors have also slashed API prices, such as OpenAI halving prices for GPT-6 Sol and GPT-6 Luna.
Where it is available: The spending slowdown is observed among companies using the Ramp platform, impacting revenue trajectories for OpenAI and Anthropic.
Example: Developers are routing workloads to cheaper light models or using Jev, a lightweight tool for choosing between predetermined outputs, to handle simple tasks.

The headline and explanation on this page are based on TechCrunch's post. The full details are in the original.

Ramp data shows AI spending stalling just as Anthropic nears its IPO