Databricks launches AI Index tracking enterprise AI token usage and coding-agent benchmarks
Introducing the Databricks AI Index: The state of enterprise AI tokenomics
Databricks launched the Databricks AI Index, a public resource combining anonymized usage and spending data from 10,000+ companies on Unity Gateway with weekly coding-agent evaluations drawn from its own private codebase. The index shows open-weight models rising from 7.3% of enterprise token traffic in July to 30.9% in November, and enterprises using two or more model providers climbing from 28% to 64% over the same period. Benchmark results also show cost per completed task diverging sharply from token price, with GPT-6 Luna and Claude Opus 5.5 both at 89.1% task success but $0.05 versus $1.62 per task.
Databricks opens its internal model evaluation suite and anonymized enterprise token data, offering a rare view of how model and harness choices translate into real coding costs.
Source:Databricks Blog · databricks.com