Google: Gemma 3 12B Performance
Try Gemma 3 12B on us. Starter credit covers your first requests, then one key reaches all 6 provider routes with automatic failover.
google/gemma-3-12b-it
Measured performance
Continuously sampled p50/p95 time-to-first-token (TTFT), effective throughput, and success rate for Google: Gemma 3 12B. Effective throughput uses provider-reported output tokens over complete request time. Unsupported route and probe-configuration rows are separated from provider downtime, and no prompt or output content is stored.
| Provider | p50 TTFT | p95 TTFT | Effective throughput | Uptime | Config excluded | Availability samples |
|---|---|---|---|---|---|---|
| deepinfra | 1124 ms | 1124 ms | — | 100.00% | — | 1 |
| amazon-bedrock | 1152 ms | 1187 ms | — | 100.00% | — | 2 |
| databricks | 1245 ms | 2879 ms | — | 100.00% | — | 3 |
| nextbit-eu | 1459 ms | 1725 ms | 109 tok/s n=1 | 100.00% | — | 7 |
| nextbit | 1521 ms | 1580 ms | — | 100.00% | — | 3 |
Full provider & model leaderboard.
6 routes.
More routes give the auto router more room to fail over around provider 429 and 5xx responses.
Gateway overhead is measured separately.
Public status separates TLS/health overhead from full model latency so slow LLMs do not inflate the router metric.
Metadata rollups.
Status samples store latency, outcome, provider, model, route, cost, and region metadata only.
View public status or inspect provider routes.