Token and wallet share by proprietary provider
Unity Gateway Token Share - Proprietary Models
Updated Nov 2, 2026
- AnthropicFable 5.1Opus 5.5Sonnet 5.5Opus 5Sonnet 5Opus 4.8Older Anthropic Models
- Older Anthropic Models
- OpenAIGPT 6.1 SolGPT 6 AstraGPT 6 SolGPT 6 LunaGPT 5.6 SolGPT 5 NanoOlder OpenAI Models
- Older OpenAI Models
- Other proprietaryGemini, Grok
Other proprietary
- Gemini, Grok
Input plus output tokens routed through Unity Gateway, including traffic routed to external model providers.
Unity Gateway Wallet Share - Proprietary Models
Updated Nov 2, 2026
- AnthropicFable 5.1Opus 5.5Sonnet 5.5Opus 5Sonnet 5Opus 4.8Older Anthropic Models
- Older Anthropic Models
- OpenAIGPT 6.1 SolGPT 6 AstraGPT 6 SolGPT 6 LunaGPT 5.6 SolGPT 5 NanoOlder OpenAI Models
- Older OpenAI Models
- Other proprietaryGemini, Grok
Other proprietary
- Gemini, Grok
Tokens priced at each provider's published list price, with cache reads at the cache-read rate. Includes traffic routed to external model providers; not what Databricks bills.
Token and wallet share by open-weight provider
Unity Gateway Token Share - Open-Weight Models
Updated Nov 2, 2026
- ZhipuGLM 5.3 FlashGLM 5.3GLM 5.2Older Zhipu Models
- GLM 5.3 Flash
- Older Zhipu Models
- DeepSeekDS V4.1 FlashDS V4 FlashOlder DeepSeek Models
- DS V4.1 Flash
- Older DeepSeek Models
- MoonshotKimi K3Older Moonshot Models
- Older Moonshot Models
- Other open-weightOther
Other open-weight
Input plus output tokens routed through Unity Gateway, including traffic routed to external model providers.
Unity Gateway Wallet Share - Open-Weight Models
Updated Nov 2, 2026
- ZhipuGLM 5.3 FlashGLM 5.3GLM 5.2Older Zhipu Models
- GLM 5.3 Flash
- Older Zhipu Models
- DeepSeekDS V4.1 FlashDS V4 FlashOlder DeepSeek Models
- DS V4.1 Flash
- Older DeepSeek Models
- MoonshotKimi K3Older Moonshot Models
- Older Moonshot Models
- Other open-weightOther
Other open-weight
Tokens priced at each provider's published list price, with cache reads at the cache-read rate. Includes traffic routed to external model providers; not what Databricks bills.
How enterprises mix models over time
Companies Using Multiple Model Providers with Unity Gateway
Updated Nov 2, 2026
- Companies using2+ providers3+ providers4+ providers
Companies using
- 2+ providers
- 3+ providers
- 4+ providers
Among companies active each week, counts every provider a company has used since July 20, 2026. Gemini and Gemma count as one provider, as do GPT and GPT-OSS.
Unity Gateway Token Share - Proprietary vs Open-Weight
Updated Nov 2, 2026
- Model typeProprietaryOpen-weight
Input plus output tokens to Databricks-hosted models on Unity Gateway. Open-weight: GLM, Kimi, DeepSeek, Qwen, Llama, GPT-OSS and Gemma; the rest is proprietary (Claude, GPT, Gemini and Grok).
What a user spends in a week
Unity Gateway User Spend Percentiles (Weekly)
Updated Nov 2, 2026
Each user's input, cached and output tokens priced at each provider's published list price; human users only.
04/Coding Frontier
Cost/Quality Frontier for Coding
Measured on an internal benchmark built from Unity Gateway traces of real enterprise coding tasks from the Databricks engineering organization.
Cost per Task vs Quality
Updated Nov 10, 2026
Built from internal benchmarks on the Databricks codebase. Each model is shown in the harness, Claude Code or Codex, that sits closest to the frontier.
Highest Quality
Top 10 models, ranked by highest pass rate on the coding benchmark.
Shortest Time per Task
Top 10 models, ranked by shortest average time to finish a task (minutes).
Lowest Cost per Task
Top 10 models, ranked by lowest average cost per task (USD).
05/Further reading
Publications
- Nov 4, 2026The Enterprise Case for Open-Weight Models on DatabricksWhy enterprises run a mix of open-weight and frontier models, weighing cost, data control and governance.
Why enterprises run a mix of open-weight and frontier models, weighing cost, data control and governance.
- Nov 1, 2026The new economics of AI cost managementAn interview on measuring AI by the value it returns per dollar, moving from tokenmaxxing to valuemaxxing.
An interview on measuring AI by the value it returns per dollar, moving from tokenmaxxing to valuemaxxing.
- Oct 22, 2026The AI Bill is comingWhy AI coding costs rise with usage, and how engineering leaders control them with gateways, routing and visibility.
Why AI coding costs rise with usage, and how engineering leaders control them with gateways, routing and visibility.
- Sep 30, 2026How to scale agentic applications without creating AI sprawlShared infrastructure for model choice, context and control, so each new agent does not rebuild governance and cost controls.
Shared infrastructure for model choice, context and control, so each new agent does not rebuild governance and cost controls.
- Sep 28, 2026How Databricks rolls out frontier models to 12,000 employees on Day 1Databricks' internal process for releasing new frontier models company-wide, covering experimental access, per-user budgets, and evaluation.
Databricks' internal process for releasing new frontier models company-wide, covering experimental access, per-user budgets, and evaluation.
- Sep 1, 2026How we eliminated $1 million a year of wasted AI agent spend in one hourUsing Unity Gateway tracing and Genie One to find and fix bugs in MCP tool servers that caused wasted agent token spend.
Using Unity Gateway tracing and Genie One to find and fix bugs in MCP tool servers that caused wasted agent token spend.
- Aug 13, 2026Smart Routing in Unity GatewayIntroduces Smart Routing, a Unity Gateway feature that selects a model for each coding task.
Introduces Smart Routing, a Unity Gateway feature that selects a model for each coding task.
- Aug 7, 2026Managing AI Coding Costs at ScaleApproaches to controlling AI coding tool spend while keeping broad developer access.
Approaches to controlling AI coding tool spend while keeping broad developer access.
- Jul 8, 2026Benchmarking Coding Agents on Databricks' Multi-Million Line CodebaseThe internal benchmark, built from merged pull requests, behind the cost and quality frontier on this page.
The internal benchmark, built from merged pull requests, behind the cost and quality frontier on this page.