CoreWeave
21 models served.
| Model | Agent Index | Input | Output | Agent $/M | Uptime 24h | Context | Quant |
|---|---|---|---|---|---|---|---|
| DeepSeek V4.1 Flash 1017 | 57.6 | $0.30 | $1.20 | $0.10 | 100.0% | 1M | fp8 |
| GLM-5.3 | 55.2 | $0.75 | $3.00 | $0.40 | 99.7% | 1M | nvfp4 |
| DeepSeek V4.1 Flash | 51.2 | $0.13 | $0.65 | $0.050 | 100.0% | 1M | fp8 |
| GLM-5.3-Flash | 48.3 | $0.12 | $0.38 | $0.050 | 99.9% | 1M | nvfp4 |
| GLM-5.2 | 47.3 | $0.62 | $2.42 | $0.32 | 100.0% | 1M | fp4 |
| Kimi K2.6 | 43.7 | $0.65 | $3.41 | $0.38 | 99.8% | 262K | fp4 |
| Kimi K2.7 Code | 42.1 | $0.60 | $3.00 | $0.33 | 98.5% | 262K | int4 |
| MiniMax-M3 | 36.7 | $0.23 | $0.96 | $0.12 | 96.9% | 262K | fp4 |
| GPT OSS 120B | 22.1 | $0.030 | $0.17 | $0.040 | 98.9% | 131K | fp4 |
| Beam 1.0 501B A23B | — | $0.30 | $1.00 | $0.12 | 97.2% | 1M | nvfp4 |
| DeepSeek V3.1 | — | $0.50 | $1.50 | $0.30 | 100.0% | 161K | fp8 |
| DeepSeek V4 Flash 0731 | — | $0.13 | $0.27 | $0.070 | 99.9% | 262K | fp8 |
| DeepSeek V4 Pro 0813 | — | $1.31 | $3.96 | $0.41 | 100.0% | 1M | fp8 |
| Gemma 4 26B A4B IT | — | $0.10 | $0.30 | $0.070 | 100.0% | 262K | bf16 |
| Gemma 4 31B IT | — | $0.10 | $0.34 | $0.11 | 100.0% | 262K | fp4 |
| GPT OSS 20B | — | $0.030 | $0.13 | $0.030 | 99.6% | 131K | fp4 |
| Granite 4.2 8B | — | $0.060 | $0.15 | $0.030 | 98.8% | 131K | bf16 |
| Nemotron 3.5 Lightning 30B A3B | — | $0.070 | $0.20 | $0.050 | 100.0% | 262K | bf16 |
| Nemotron 3.5 Super VL | — | $0.15 | $0.47 | $0.060 | 100.0% | 1M | nvfp4 |
| Qwen3.6 35B-A3B | — | $0.10 | $0.89 | $0.090 | 100.0% | 262K | fp8 |
| Qwen3.8 27B | — | $0.40 | $3.00 | $0.32 | 99.4% | 262K | fp8 |