AI Token Prices Drop as Usage Volume Soars, Data Shows
AI token volume surged 27.2% month-over-month while API spending declined from recent peaks, driven by price competition between major foundation model providers.

Data from corporate spend trackers and API aggregators shows that while overall artificial intelligence consumption is reaching record highs, token pricing drops are causing revenue growth to decelerate. Token spending dropped 6.3 percent from its peak recorded two weeks earlier, even as total monthly token volume jumped 27.2 percent. Month over month, total spending grew 12 percent, revealing a widening gap between API usage volume and provider revenues.
Declining token prices drive a wedge between usage and spend
According to data compiled by Ramp, the average cost per million tokens for OpenAI and Anthropic combined dropped to $0.45 in mid-November, reflecting an 11.2 percent decrease from the previous month. Model pricing shows OpenAI averaging $0.38 per million tokens, compared to $0.50 per million tokens for Anthropic. Intense price competition between OpenAI, Anthropic, and Chinese vendors has continuously pushed down token costs across the board.
Data from API routing platform OpenRouter mirrors this trend. An analysis published on X by Peter Walker shows that weekly token volume for closed-weight models from Anthropic, OpenAI, and Google rose 1,215 percent since the beginning of 2026. Over the exact same timeframe, weekly spending grew far more slowly at 455 percent. OpenRouter data through the week ending November 9 indicates that weekly spending peaked during the week of October 5 before turning downward.
Market shares and corporate adoption rates
In the week ending November 15, 2026, Anthropic accounted for 58.7 percent of total token spend—up 1.7 percentage points from October—and held 59.4 percent of token volume, according to the Ramp AI Index. OpenAI followed with 37.2 percent of spending and 37.7 percent of volume. SpaceXAI (xAI) captured 2.4 percent of spending, while Google represented 1.4 percent. On X, Kharazian noted Anthropic maintained this commanding position on its IPO day. During October, Anthropic's token spending grew 0.5 percent month over month, whereas OpenAI's spend grew 2.1 percent.
Corporate adoption also continues to expand. In October 2026, 56.8 percent of tracked US businesses paid for AI services or products, up 1 percentage point from September and 13.3 percentage points higher than the previous year. Anthropic led corporate adoption with 44.8 percent of firms paying for its products, compared to 40.6 percent for OpenAI.
Revenue growth cools on the provider side
Payment platform Stripe published its own AI Index measuring incoming payments across roughly 20,000 AI vendors. Stripe reported that revenue across tracked AI companies increased 103 percent over the 52 weeks ending November 8, significantly outperforming the 33 percent growth recorded in the SaaS Index. Payment activity in the US grew 104 percent year-over-year, followed by the UK at 83 percent, the rest of the world at 76 percent, and the EU at 67 percent.
Despite high annual growth, Stripe's metrics reveal clear signs of slowing momentum. The median AI startup grew just 41 percent over the 52-week period, demonstrating that top-tier providers drive most of the aggregate gains. Additionally, annualized growth calculated over a 13-week period fell to approximately 80 percent, down from 140 percent recorded earlier in the year.
What it means for developers
For developers building software on LLM infrastructure, the sharp decline in token pricing significantly reduces the cost of running large-scale workloads and complex AI workflows. Reduced API costs enable teams to run higher token throughput, batch processing, and agentic workflows without proportional budget increases.
To manage costs and access multiple model options seamlessly, developers can try top AI models cheaply through one API at https://apixoai.online. As token margins compress, foundation model providers will likely focus on improving model reliability, workflow integrations, and advanced capabilities to maintain revenue momentum in an increasingly competitive market.
Source: AI industry growth remains strong, but the pace is slowing — The Decoder. Written by the Apixo team from that report.
One key for Claude, GPT, GLM, DeepSeek and more. Pay per token with crypto.
Get your API key

