OpenAIAPI Host BenchmarksEstimated
GPT-oss-120B — 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
3000 t/s
Lowest latency (TTFT)
90 ms P50
Best price
$0.05 / $0.2
Blended Pricing Workload Calculator
Switch between standard conversational ratios, agentic caching loops, and heavy RAG workloads.
| Provider & Host | Hardware Stack | Output Speed (TPS) | Latency TTFT (P50) | Input Price ($/1M) | Output Price ($/1M) | Cache Read ($/1M) | Blended Price ($/1M) | SLA |
|---|---|---|---|---|---|---|---|---|
OpenAIOfficial | Cloud TPU / Nvidia H100 | 240 t/s | 800 ms (1.12 s p90) | $0.1 | $0.4 | $0.01 | $0.18 | 99.99% |
| Groq LPU v2 (SRAM) | 1008 t/s | 90 ms (120 ms p90) | $0.09 | $0.34 | $0.008 | $0.15 | 99.95% | |
| CS-3 Wafer-Scale Engine | 3000 t/s | 120 ms (160 ms p90) | $0.09 | $0.36 | — | $0.16 | 99.90% | |
| Nvidia H100 SXM5 | 360 t/s | 640 ms (880 ms p90) | $0.07 | $0.28 | $0.006 | $0.12 | 99.95% | |
| Nvidia H100 / L40S | 312 t/s | 720 ms (1.00 s p90) | $0.05 | $0.2 | $0.005 | $0.09 | 99.90% | |
| Nvidia B200 / H100 | 456 t/s | 560 ms (760 ms p90) | $0.06 | $0.26 | $0.006 | $0.11 | 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.