XiaomiAPI Host BenchmarksEstimated
MiMo V2.6 Pro — 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
900 t/s
Lowest latency (TTFT)
90 ms P50
Best price
$0.22 / $0.47
Blended Pricing Workload Calculator
Switch between standard conversational ratios, agentic caching loops, and heavy RAG workloads.
| Provider & Host | Hardware Stack | Output Speed (TPS) | Latency TTFT (P50) | Input Price ($/1M) | Output Price ($/1M) | Cache Read ($/1M) | Blended Price ($/1M) | SLA |
|---|---|---|---|---|---|---|---|---|
XiaomiOfficial | Cloud TPU / Nvidia H100 | 72 t/s | 340 ms (480 ms p90) | $0.48 | $0.95 | $0.048 | $0.6 | 99.99% |
| Groq LPU v2 (SRAM) | 302 t/s | 90 ms (120 ms p90) | $0.41 | $0.81 | $0.038 | $0.51 | 99.95% | |
| CS-3 Wafer-Scale Engine | 900 t/s | 120 ms (160 ms p90) | $0.43 | $0.85 | — | $0.54 | 99.90% | |
| Nvidia H100 SXM5 | 108 t/s | 270 ms (370 ms p90) | $0.31 | $0.66 | $0.029 | $0.4 | 99.95% | |
| Nvidia H100 / L40S | 94 t/s | 310 ms (430 ms p90) | $0.22 | $0.47 | $0.024 | $0.28 | 99.90% | |
| Nvidia B200 / H100 | 137 t/s | 240 ms (320 ms p90) | $0.29 | $0.62 | $0.029 | $0.37 | 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.