Databricks
Explore Databricks models on TrustedRouter with current routes, token pricing, policy sources, privacy posture, regional availability, and API support.
Databricksdatabricks
No provider claim
| Provider | Databricks |
|---|---|
| Provider website | https://www.databricks.com/ |
| Models | 11 public models |
| Prepaid routes | 11 |
| BYOK routes | 0 |
| Zero data retention | no |
| Confidential compute | no |
| Provider E2EE | no |
| Policy note | Databricks Foundation Model APIs are a Designated Service and pay-per-token workloads are HIPAA compliant. Databricks may temporarily process or store inputs and outputs for abuse prevention, and may process data outside the originating cloud or region. This route is therefore standard privacy, not ZDR, confidential compute, or end-to-end encrypted. Policy source |
Measured performance
31 samplesContinuously sampled across Databricks's routed models: p50 TTFT, effective throughput, and success rate. Effective throughput uses provider-reported output tokens over complete request time. Unsupported route and probe-configuration rows are separated from provider downtime. No prompt or output content stored.
| p50 TTFT | 2252 ms |
|---|---|
| Effective throughput | 109 tok/s n=1 |
| Uptime | 100.00% |
| Model | p50 TTFT | p50 TTFB | Effective throughput | Uptime | Config excluded | Availability samples |
|---|---|---|---|---|---|---|
| meta-llama/llama-3.1-8b-instruct | 1587 ms | 1587 ms | — | 100.00% | — | 4 |
| google/gemma-3-12b-it | 1869 ms | 1869 ms | — | 100.00% | — | 6 |
| openai/gpt-oss-120b | 2026 ms | 2026 ms | 109 tok/s n=1 | 100.00% | — | 5 |
| qwen/qwen3-next-80b-a3b-instruct | 2252 ms | 2252 ms | — | 100.00% | — | 3 |
| meta-llama/llama-4-maverick | 3163 ms | 3162 ms | — | 100.00% | — | 6 |
| meta-llama/llama-3.3-70b-instruct | 3179 ms | 3179 ms | — | 100.00% | — | 5 |
| openai/gpt-oss-20b | 3540 ms | 3540 ms | — | 100.00% | — | 2 |
Databricks performance history · Full provider & model leaderboard.
Models served by Databricks.
Each row links to pricing, provider, benchmark, and API pages for the model.
| Model | AI IQ | Context | Endpoints | Prompt | Completion | Routes |
|---|---|---|---|---|---|---|
google/gemma-3-12b-itGoogle: Gemma 3 12B |
— | 131,072 | 1 | $0.1575/1M | $0.525/1M | prepaid |
meta-llama/llama-3.1-8b-instructMeta: Llama 3.1 8B Instruct |
— | 131,072 | 1 | $0.1575/1M | $0.4725/1M | prepaid |
meta-llama/llama-3.3-70b-instructMeta: Llama 3.3 70B Instruct |
— | 131,072 | 1 | $0.525/1M | $1.575/1M | prepaid |
meta-llama/llama-4-maverickMeta: Llama 4 Maverick |
IQ 90#91 | 1,048,576 | 1 | $0.525/1M | $1.575/1M | prepaid |
moonshotai/kimi-k3MoonshotAI: Kimi K3 |
IQ 123#14 | 1,048,576 | 1 | $3.15/1M | $15.75/1M | prepaid |
openai/gpt-oss-120bOpenAI: gpt-oss-120b |
IQ 105#53 | 131,072 | 1 | $0.1575/1M | $0.63/1M | prepaid |
openai/gpt-oss-20bOpenAI: gpt-oss-20b |
IQ 100#70 | 131,072 | 1 | $0.0735/1M | $0.315/1M | prepaid |
qwen/qwen3-next-80b-a3b-instructQwen: Qwen3 Next 80B A3B Instruct |
— | 262,144 | 1 | $0.1575/1M | $1.26/1M | prepaid |
qwen/qwen3.5-122b-a10bQwen: Qwen3.5-122B-A10B |
— | 262,144 | 1 | $0.231/1M | $2.31/1M | prepaid |
thinkingmachines/inklingThinking Machines: Inkling |
IQ 108#47 | 1,048,576 | 1 | $1.05/1M | $4.2525/1M | prepaid |
z-ai/glm-5.2Z.ai: GLM 5.2 |
IQ 120#16 | 1,048,576 | 1 | $1.47/1M | $4.62/1M | prepaid |
Questions
Does Databricks have zero data retention?
TrustedRouter does not currently mark Databricks as provider-level zero data retention. Use trustedrouter/zdr or provider.min_privacy=zdr to select a different eligible route, and review the linked policy source for changes.
Is Databricks end-to-end encrypted?
TrustedRouter does not currently mark Databricks as end-to-end encrypted at the provider boundary. The TrustedRouter gateway is still attested, but the selected provider normally receives the request in order to run the model. Use trustedrouter/e2e for the stronger route requirement.
Which Databricks models are available through TrustedRouter?
This page currently lists 11 public Databricks models, with live pricing, route count, context length, measured performance when available, and links to each model's provider and benchmark pages.