AsyncAPI Host Benchmarks
Async Pro V1 0 — API Providers & Speed Matrix
Compare live throughput, TTFT latency distributions, prompt cache pricing, and hardware stacks across cloud hosting endpoints.
Fastest Provider
750 t/s
Lowest Latency (TTFT)
0.09s P50
Best Price Value
$0.45 / $1.5
Blended Pricing Workload Calculator
Switch between standard conversational ratios, agentic caching loops, and heavy RAG workloads.
| Provider & Host | Hardware Stack | Output Speed (TPS) | Latency TTFT (P50) | Input Price ($/1M) | Output Price ($/1M) | Cache Read ($/1M) | Blended Price ($/1M) | SLA |
|---|---|---|---|---|---|---|---|---|
AsyncOfficial | Cloud TPU / Nvidia H100 | 60 t/s | 0.8s (1.12s p90) | $1 | $3 | $0.1 | $1.5 | 99.99% |
Groq Cloud | Groq LPU v2 (SRAM) | 252 t/s | 0.09s (0.12s p90) | $0.85 | $2.55 | $0.08 | $1.27 | 99.95% |
Cerebras Cloud | CS-3 Wafer-Scale Engine | 750 t/s | 0.12s (0.16s p90) | $0.9 | $2.7 | — | $1.35 | 99.90% |
Together AI | Nvidia H100 SXM5 | 90 t/s | 0.64s (0.88s p90) | $0.65 | $2.1 | $0.06 | $1.01 | 99.95% |
DeepInfra | Nvidia H100 / L40S | 78 t/s | 0.72s (1s p90) | $0.45 | $1.5 | $0.05 | $0.71 | 99.90% |
Fireworks AI | Nvidia B200 / H100 | 114 t/s | 0.56s (0.76s p90) | $0.6 | $1.95 | $0.06 | $0.94 | 99.99% |