GoogleAPI Host BenchmarksEstimated
Gemma 4 31B — API Providers & Speed Matrix
Modelled throughput, TTFT latency and prompt-cache pricing estimates across cloud hosting endpoints, derived from official list prices and public benchmarks.
Fastest provider
438 t/s
Lowest latency (TTFT)
90 ms P50
Best price
$0.03 / $0.15
Blended Pricing Workload Calculator
Switch between standard conversational ratios, agentic caching loops, and heavy RAG workloads.
| Provider & Host | Hardware Stack | Output Speed (TPS) | Latency TTFT (P50) | Input Price ($/1M) | Output Price ($/1M) | Cache Read ($/1M) | Blended Price ($/1M) | SLA |
|---|---|---|---|---|---|---|---|---|
GoogleOfficial | Cloud TPU / Nvidia H100 | 35 t/s | 1.10 s (1.54 s p90) | $0.075 | $0.3 | $0.007 | $0.13 | 99.99% |
| Groq LPU v2 (SRAM) | 147 t/s | 90 ms (120 ms p90) | $0.06 | $0.26 | $0.006 | $0.11 | 99.95% | |
| CS-3 Wafer-Scale Engine | 438 t/s | 120 ms (160 ms p90) | $0.07 | $0.27 | — | $0.12 | 99.90% | |
| Nvidia H100 SXM5 | 53 t/s | 880 ms (1.21 s p90) | $0.05 | $0.21 | $0.004 | $0.09 | 99.95% | |
| Nvidia H100 / L40S | 46 t/s | 990 ms (1.38 s p90) | $0.03 | $0.15 | $0.004 | $0.06 | 99.90% | |
| Nvidia B200 / H100 | 67 t/s | 770 ms (1.04 s p90) | $0.04 | $0.2 | $0.004 | $0.08 | 99.99% |
Figures are modelled estimates based on published list prices and public throughput benchmarks — not measured on live endpoints. Verify current pricing and terms with each host before committing.