On November 16 Databricks launched the Databricks AI Index on its official blog and X, combining anonymized usage from 10,000+ companies on Unity Gateway with coding-agent evaluations built on real internal pull requests and hidden tests, updated weekly. Open-weight models grew from 7.3% of enterprise tokens in July to 30.9% in November, and the share of enterprises using a single model provider fell from about 64% to 28%. Claude Opus 5.5 rose from 7.2% to 37.9% of proprietary-model tokens in six weeks. In the coding benchmark, GPT-6 Luna and Opus 5.5 both reached an 89.1% success rate at about $0.05 and $1.62 per task respectively.

Key Takeaways

  • ✓Data: anonymized usage from 10,000+ companies on Unity Gateway plus coding-agent evaluations built on real pull requests from Databricks' private codebase with hidden tests; current data covers September 21 to November 7, updated weekly
  • ✓Open-weight token share went from 7.3% in July to 30.9% in November; among open-weight models, GLM 5.3 Flash had 33.3%, DeepSeek V4.1 Flash 28.0% and GLM 5.3 21.6% in the week of November 2
  • ✓Opus 5.5 rose from 7.2% to 37.9% of proprietary tokens in six weeks, lifting Anthropic from 57.6% to 68.8%; Fable 5.1 fell from 5.5% to 0.6%
  • ✓Coding benchmark: Fable 5.1 leads at 96.0% ($7.47 per task), Muse Wonder 94.9% ($1.53); DeepSeek V4.1 and GLM 5.3 tie for third at 93.1%
  • ✓Token price is not cost: Gemini 3.8 Flash averaged 140 turns per task at about $11.97 per task; median weekly user spend was $2.49 versus $699.74 at the 99th percentile
Databricks Launches Public AI Index: Open-Weight Models Rise From 7.3% to 30.9% of Enterprise Tokens, Opus 5.5 Takes 37.9% of Proprietary Tokens in Six Weeks, GPT-6 Luna Matches Opus 5.5 at About 32x Lower Cost per Coding Task
🖼️Official Media
Click to view high-res
🧭

Turn your technical choice into a development budget

Compare 40 dev plans & simulate token costs vs $20/mo subscriptions

🔬

In-Depth Technical Analysis

Background

Databricks says a new flagship model has launched roughly every week since May, making it hard for enterprises to know which model to use, what it costs and when to switch. On November 16 it launched the public Databricks AI Index on its official blog and X, making its internal evaluation suite permanently public with weekly updates.

How it works

The Index has two data sources. One is anonymized usage and spend from 10,000+ companies on Unity Gateway, including traffic to external providers such as Anthropic and OpenAI. The other is coding-agent evaluations built on real pull requests, investigations and development tasks from Databricks' private codebase, scored with hidden tests rather than an LLM judge, and comparing harnesses such as Claude Code and Codex.

Numbers that matter

Open-weight token share rose from 7.3% to 30.9%, and Opus 5.5 went from 7.2% to 37.9% of proprietary tokens in six weeks. In the coding benchmark, Fable 5.1 leads at 96.0% but costs $7.47 per task, while Muse Wonder scores 94.9% at $1.53. DeepSeek V4.1 and GLM 5.3 tie for third at 93.1%, and GPT-6 Luna and Opus 5.5 both hit 89.1% at $0.05 versus $1.62 per task.

What it means for developers

The Index is free at developers.databricks.com/ai-index. When choosing a coding agent, compare cost per completed task rather than token price: Gemini 3.8 Flash is cheap per token but averaged 140 turns and about $11.97 per task. Replay your own traces to compare model and harness combinations, and route routine work to cheaper models.

Action HubReady to adopt this in production?

Benchmark side-by-side against alternatives, or calculate monthly token cost vs subscription break-even.

Up NextArtificial Analysis Coding Agent Index v2.1 Moves to FrontierCode 2.0 and Terminal-Bench 4.1: Claude Fable 5.2 and Claude Opus 5.5 Tie for First at 58, Scores Drop 5 Points on Average
Read Now→