Databricks AI Index: enterprises stick with "good enough" models over newer releases
pwendell · x · 2026-11-17
Databricks launched an AI Index built on aggregated, anonymized data from thousands of enterprises via Unity Gateway, showing token and wallet share by provider: proprietary models split between Anthropic (Fable 5.1, Opus 5.5, Sonnet 5.5, etc.) and OpenAI (GPT 6.1 Sol, GPT 6 Astra, etc.), and open-weight models led by Zhipu (GLM 5.3 Flash/5.3/5.2), DeepSeek (V4.1 Flash, V4 Flash) and Moonshot (Kimi K3), priced at published list rates with cache reads at cache rates.
Commenting on why Astra/Fable usage appears low in the new dataset, pwendell says the #1 reason customers cite is that existing Opus/Sol/OSS models have simply gotten "good enough" for a huge fraction of enterprise workloads, dampening upgrade momentum.
The dashboard also ranks coding agents on quality-per-cost, making it a real-world barometer of enterprise model adoption.
More from Companies & People
- Ex-OpenAI employee says external safety assessors got far less access than Altman promised — j_asminewang · 2026-11-17
- Stanford HAI awards $3.5M to seven teams via 2026 Hoffman-Yee Research Grants — StanfordHAI · 2026-11-17
- Runway CTO Kamil Sindi to break down training and real-time serving of video foundation models at AI Dev NYC — DeepLearningAI · 2026-11-17
- Kai-Fu Lee pushes CEOs to hire AI agents as Chief of Staff—or even Co-CEO — kaifulee · 2026-11-17
- Iliad January Fellowship Applications Close Today: Paid Math-Based AI Alignment Research — austinc3301 · 2026-11-17
- Omar Khattab on impactful AI research: invest in projects, not papers — lateinteraction · 2026-11-17