DeepInfra
Models on DeepInfra · prices & measurements for models served by DeepInfra’s APIUSD / MTOK · VERIFIED 4 MINUTES AGO
| MODEL | INPUT | OUTPUT | CONTEXT | QUANT | TTFT | TPS | SMOKE | VS BEST ELSEWHERE |
|---|---|---|---|---|---|---|---|---|
| $0.019 | $0.03 | 131K | fp8 | ✕ | ✕ | — | cheapest here | |
| $0.02 | $0.04 | 131K | fp8 | 1.1s° | 37° tok/s | 3/5 | cheapest here | |
| $0.03 | $0.14 | 131K | bf16 | 0.26s° | 141° tok/s | 2/5 | cheapest here | |
| $0.037 | $0.17 | 131K | bf16 | 0.55s° | 54° tok/s | 4/5 | $0.03 elsewhere | |
| SALlama 3 8B LunarisSao10KBEST PRICE | $0.04 | $0.05 | 8K | fp8 | 0.15s° | 88° tok/s | — | cheapest here |
| $0.05 | $0.20 | 262K | fp4 | 0.83s° | 441° tok/s | — | cheapest here | |
| $0.05 | $0.10 | 131K | bf16 | 0.62s° | 72° tok/s | — | only provider | |
| $0.05 | $0.15 | 131K | bf16 | 0.64s° | 42° tok/s | — | only provider | |
| $0.05 | $0.08 | 33K | fp8 | 1.4s° | 51° tok/s | — | only provider | |
| INLing-3.0-flashInclusionAI | $0.06 | $0.18 | 131K | bf16 | 0.84s° | 83° tok/s | — | $0.021 elsewhere |
| $0.06 | $0.40 | 203K | bf16 | 0.55s° | 98° tok/s | — | cheapest here | |
| $0.07 | $0.34 | 262K | fp8 | 0.65s° | 25° tok/s | 5/5 | $0.042 elsewhere | |
| MIPhi 4Microsoft | $0.07 | $0.14 | 16K | bf16 | 0.71s° | 84° tok/s | — | only provider |
| $0.075 | $0.20 | 128K | fp8 | 0.9s° | 59° tok/s | — | cheapest here | |
| $0.08 | $0.20 | 262K | bf16 | 0.29s° | 103° tok/s | — | cheapest here | |
| $0.08 | $0.18 | 1.05M | fp8 | 0.56s° | 86° tok/s | — | $0.04 elsewhere | |
| $0.08 | $0.28 | 41K | fp8 | 2.4s° | 25° tok/s | — | cheapest here | |
| $0.08 | $0.16 | 131K | fp8 | 5.5s° | 14° tok/s | 3/5 | cheapest here | |
| $0.085 | $0.40 | 262K | bf16 | 2.2s° | — | — | cheapest here | |
| $0.09 | $0.18 | 1.05M | fp8 | 1.9s° | 87° tok/s | 5/5 | $0.056 elsewhere | |
| $0.09 | $0.34 | 262K | fp4 | 0.84s° | 56° tok/s | 5/5 | $0.08 elsewhere | |
| $0.09 | $1.10 | 262K | fp8 | 0.58s° | 104° tok/s | — | cheapest here | |
| $0.09 | $0.55 | 262K | fp8 | 0.41s° | 31° tok/s | 5/5 | $0.087 elsewhere | |
| $0.10 | $0.95 | 262K | fp8 | 0.54s° | 159° tok/s | 2/5 | $0.07 elsewhere | |
| $0.10 | $0.15 | 262K | bf16 | 0.68s° | 58° tok/s | — | cheapest here | |
| $0.10 | $0.30 | 328K | fp8 | 0.41s° | 50° tok/s | — | cheapest here | |
| $0.10 | $0.32 | 131K | fp8 | 1.9s° | 12° tok/s | 5/5 | cheapest here | |
| $0.12 | $0.50 | 41K | fp8 | 0.53s° | 216° tok/s | — | cheapest here | |
| $0.12 | $0.24 | 41K | fp8 | 0.82s° | 70° tok/s | — | $0.10 elsewhere | |
| TEHy3Tencent | $0.14 | $0.58 | 262K | fp8 | 0.8s° | 158° tok/s | 2/5 | $0.126 elsewhere |
| $0.14 | $1.00 | 262K | fp8 | 0.4s° | 184° tok/s | 3/5 | cheapest here | |
| $0.15 | $0.60 | 262K | fp8 | 0.61s° | 45° tok/s | — | $0.13 elsewhere | |
| $0.18 | $0.18 | 164K | bf16 | 1.4s° | 32° tok/s | — | cheapest here | |
| STStep 3.7 FlashStepFunBEST PRICE | $0.20 | $1.15 | 262K | — | 0.55s° | 148° tok/s | — | cheapest here |
| $0.20 | $0.88 | 262K | fp8 | 0.66s° | 21° tok/s | — | cheapest here | |
| $0.20 | $0.80 | 1.05M | fp8 | 0.37s° | 95° tok/s | 4/5 | cheapest here | |
| $0.24 | $0.90 | 164K | fp4 | 0.85s° | 32° tok/s | — | cheapest here | |
| $0.25 | $1.00 | 197K | fp8 | 25.6s° | 17° tok/s | 2/5 | $0.24 elsewhere | |
| $0.25 | $0.95 | 164K | fp4 | 1.5s° | 9° tok/s | 5/5 | cheapest here | |
| $0.26 | $2.60 | 262K | fp8 | 0.8s° | 81° tok/s | 3/5 | $0.195 elsewhere | |
| $0.26 | $0.38 | 164K | fp4 | 1.1s° | 10° tok/s | 5/5 | $0.209 elsewhere | |
| $0.28 | $1.10 | 524K | fp8 | 0.48s° | 19° tok/s | 5/5 | $0.23 elsewhere | |
| $0.29 | $2.40 | 262K | fp4 | 0.75s° | 123° tok/s | 3/5 | $0.26 elsewhere | |
| $0.30 | $1.20 | 131K | bf16 | 0.86s° | 177° tok/s | — | cheapest here | |
| $0.30 | $1.00 | 262K | fp4 | 0.41s° | 145° tok/s | — | $0.22 elsewhere | |
| $0.32 | $3.20 | 262K | fp8 | 0.96s° | 41° tok/s | 2/5 | $0.30 elsewhere | |
| $0.32 | $0.89 | 164K | fp4 | 0.4s° | 126° tok/s | — | $0.257 elsewhere | |
| $0.36 | $0.40 | 33K | fp8 | 1.2s° | 47° tok/s | — | cheapest here | |
| XIMiMo-V2.5Xiaomi | $0.40 | $2.00 | 262K | bf16 | 0.71s° | 39° tok/s | — | $0.119 elsewhere |
| $0.40 | $1.75 | 203K | fp4 | 0.84s° | 29° tok/s | 2/5 | cheapest here | |
| $0.40 | $0.40 | 131K | fp8 | 0.95s° | 52° tok/s | — | cheapest here | |
| GRMythoMax 13BGryphe | $0.40 | $0.40 | 4K | fp16 | 0.28s° | 53° tok/s | — | $0.06 elsewhere |
| TMInkling SmallThinking MachinesBEST PRICE | $0.45 | $1.20 | 524K | fp8 | 0.59s° | 301° tok/s | — | cheapest here |
| $0.45 | $3.00 | 262K | fp8 | 1.3s° | 25° tok/s | 3/5 | $0.39 elsewhere | |
| $0.45 | $2.25 | 262K | fp4 | 0.73s° | 43° tok/s | 3/5 | cheapest here | |
| $0.50 | $2.20 | 262K | fp4 | 3.4s° | — | — | cheapest here | |
| $0.50 | $2.00 | 203K | fp4 | 0.67s° | 78° tok/s | 3/5 | $0.43 elsewhere | |
| $0.50 | $2.15 | 164K | fp4 | 0.68s° | 20° tok/s | — | cheapest here | |
| $0.60 | $2.08 | 203K | fp4 | 0.55s° | 90° tok/s | 2/5 | cheapest here | |
| $0.68 | $3.40 | 262K | fp4 | 0.4s° | 58° tok/s | 5/5 | $0.67 elsewhere | |
| NRHermes 3 70B InstructNous Research | $0.70 | $0.70 | 131K | fp8 | 2.2s° | 39° tok/s | — | only provider |
| $0.75 | $2.40 | 1.05M | fp4 | 0.94s° | 74° tok/s | — | $0.50 elsewhere | |
| $0.75 | $3.50 | 262K | fp4 | 1.1s° | 32° tok/s | — | $0.58 elsewhere | |
| SALlama 3.1 Euryale 70B v2.2Sao10KBEST PRICE | $0.85 | $0.85 | 131K | fp8 | 0.91s° | 47° tok/s | — | cheapest here |
| TMInklingThinking MachinesBEST PRICE | $0.95 | $4.05 | 524K | fp8 | 0.46s° | 199° tok/s | — | cheapest here |
| XIMiMo-V2.5-ProXiaomi | $1.00 | $3.00 | 1.05M | fp8 | 0.58s° | 80° tok/s | — | $0.304 elsewhere |
| NRHermes 3 405B InstructNous Research | $1.00 | $1.00 | 131K | fp8 | 2.3s° | 16° tok/s | — | only provider |
| $1.05 | $3.50 | 203K | fp4 | 1s° | 13° tok/s | 2/5 | $0.91 elsewhere | |
| $1.30 | $2.60 | 1.05M | fp8 | — | — | — | $0.66 elsewhere | |
| $1.30 | $2.60 | 1.05M | fp8 | 1.3s° | 57° tok/s | — | $0.522 elsewhere | |
| $2.00 | $6.00 | 262K | fp4 | 1.3s° | 76° tok/s | 4/5 | cheapest here | |
| $2.85 | $14.25 | 1.05M | bf16 | 2s° | 25° tok/s | 4/5 | $2.60 elsewhere |
TTFT/TPS: ModelIndex-measured medians (P50, trailing 72h) for DeepInfra’s endpoints · ° = measured via OpenRouter routing · ✕ = probed, no measurable stream · SMOKE: of 5 capability probes passed · methodology on each model page