CW
CoreWeave
Inference providerdocs.coreweave.com ↗
Models on CoreWeave · prices & measurements for models served by CoreWeave’s APIUSD / MTOK · VERIFIED 4 MINUTES AGO
| MODEL | INPUT | OUTPUT | CONTEXT | QUANT | TTFT | TPS | SMOKE | VS BEST ELSEWHERE |
|---|---|---|---|---|---|---|---|---|
 gpt-oss-120bOpenAIBEST PRICE | $0.03 | $0.17 | 131K | fp4 | 0.76s° | 51° tok/s | 3/5 | cheapest here |
| gpt-oss-20bOpenAIBEST PRICE | $0.03 | $0.13 | 131K | fp4 | 0.32s° | 111° tok/s | 5/5 | cheapest here |
| IGGranite 4.1 8BIBM Granite | $0.05 | $0.10 | 131K | bf16 | 0.41s° | 118° tok/s | — | only provider |
 Nemotron 3.5 LightningNVIDIA | $0.10 | $0.25 | 262K | bf16 | 0.53s° | 397° tok/s | — | $0.08 elsewhere |
 Gemma 4 31BGoogle | $0.10 | $0.34 | 262K | bf16 | 1.6s° | 38° tok/s | 5/5 | $0.08 elsewhere |
 Qwen3 30B A3B Instruct 2507Qwen | $0.10 | $0.30 | 262K | bf16 | 0.38s° | 172° tok/s | — | $0.048 elsewhere |
 DeepSeek V4 Flash 0731DeepSeek | $0.13 | $0.28 | 262K | fp8 | 1.1s° | 230° tok/s | — | $0.04 elsewhere |
| DeepSeek V4 Flash 0423DeepSeek | $0.14 | $0.28 | 1.05M | fp8 | 1s° | 63° tok/s | — | $0.056 elsewhere |
 Llama 3.1 8B InstructMeta | $0.22 | $0.22 | 128K | bf16 | 0.42s° | 147° tok/s | 4/5 | $0.02 elsewhere |
 MiniMax M3MiniMaxBEST PRICE | $0.23 | $0.96 | 262K | fp4 | 0.93s° | 98° tok/s | 5/5 | cheapest here |
| Qwen3.6 35B A3BQwen | $0.25 | $1.25 | 262K | fp8 | 0.48s° | 197° tok/s | — | $0.07 elsewhere |
| Qwen3.5-35B-A3BQwen | $0.25 | $1.25 | 262K | fp8 | 0.49s° | 244° tok/s | 3/5 | $0.14 elsewhere |
| Qwen3.8 27BQwen | $0.40 | $3.00 | 262K | fp8 | 1.7s° | 50° tok/s | 4/5 | $0.35 elsewhere |
| DeepSeek V3.1DeepSeek | $0.55 | $1.65 | 161K | fp8 | 1.1s° | 57° tok/s | 5/5 | $0.25 elsewhere |
| Qwen3.6 27BQwen | $0.60 | $3.60 | 262K | fp8 | 1.6s° | 66° tok/s | — | $0.30 elsewhere |
 Kimi K2.6Moonshot AI | $0.65 | $3.41 | 262K | fp4 | 0.55s° | 199° tok/s | 3/5 | $0.58 elsewhere |
| Kimi K2.7 CodeMoonshot AI | $0.71 | $3.50 | 262K | int4 | 0.61s° | 196° tok/s | 5/5 | $0.67 elsewhere |
| Llama 3.3 70B InstructMeta | $0.71 | $0.71 | 128K | fp16 | 0.59s° | 86° tok/s | — | $0.10 elsewhere |
 GLM 5.2Z.ai | $0.76 | $2.42 | 262K | fp4 | ✕ | ✕ | — | $0.50 elsewhere |
| Llama 3.1 70B InstructMeta | $0.80 | $0.80 | 128K | bf16 | 0.61s° | 76° tok/s | — | $0.40 elsewhere |
| DeepSeek V4 Pro 0423DeepSeek | $1.15 | $2.55 | 1.05M | fp8 | ✕ | ✕ | 0/5 | $0.522 elsewhere |
TTFT/TPS: ModelIndex-measured medians (P50, trailing 72h) for CoreWeave’s endpoints · ° = measured via OpenRouter routing · ✕ = probed, no measurable stream · SMOKE: of 5 capability probes passed · methodology on each model page