Thinking MachinesAPI Host Benchmarks
Inkling Small — API Providers & Speed Matrix
Compare live throughput, TTFT latency distributions, prompt cache pricing, and hardware stacks across cloud hosting endpoints.
Fastest Provider
1625 t/s
Lowest Latency (TTFT)
0.09s P50
Best Price Value
$0.14 / $0.6
Blended Pricing Workload Calculator
Switch between standard conversational ratios, agentic caching loops, and heavy RAG workloads.
| Provider & Host | Hardware Stack | Output Speed (TPS) | Latency TTFT (P50) | Input Price ($/1M) | Output Price ($/1M) | Cache Read ($/1M) | Blended Price ($/1M) | SLA |
|---|---|---|---|---|---|---|---|---|
Thinking MachinesOfficial | Cloud TPU / Nvidia H100 | 130 t/s | 1.58s (2.21s p90) | $0.3 | $1.2 | $0.03 | $0.52 | 99.99% |
Groq Cloud | Groq LPU v2 (SRAM) | 546 t/s | 0.09s (0.12s p90) | $0.26 | $1.02 | $0.024 | $0.45 | 99.95% |
Cerebras Cloud | CS-3 Wafer-Scale Engine | 1625 t/s | 0.12s (0.16s p90) | $0.27 | $1.08 | — | $0.47 | 99.90% |
Together AI | Nvidia H100 SXM5 | 195 t/s | 1.26s (1.74s p90) | $0.2 | $0.84 | $0.018 | $0.36 | 99.95% |
DeepInfra | Nvidia H100 / L40S | 169 t/s | 1.42s (1.98s p90) | $0.14 | $0.6 | $0.015 | $0.26 | 99.90% |
Fireworks AI | Nvidia B200 / H100 | 247 t/s | 1.11s (1.5s p90) | $0.18 | $0.78 | $0.018 | $0.33 | 99.99% |