Offerings · 11 models · 190 configs

Qwen2.5 7B Instruct · 7.6B · 131,072 ctx

GPU×QuantWeights GBVRAM GB$/hr
NVIDIA GeForce RTX 3090 1 fp16 16 24 $0.20
NVIDIA GeForce RTX 3090 1 fp8 8 24 $0.20
NVIDIA GeForce RTX 4090 1 fp16 16 24 $0.35
NVIDIA GeForce RTX 4090 1 fp8 8 24 $0.35
NVIDIA RTX A6000 1 fp16 16 48 $0.45
NVIDIA RTX A6000 1 fp8 8 48 $0.45
NVIDIA L40S 1 fp16 16 48 $0.55
NVIDIA L40S 1 fp8 8 48 $0.55
NVIDIA A100 80GB 1 fp16 16 80 $1.27
NVIDIA A100 80GB 1 fp8 8 80 $1.27
NVIDIA H100 PCIe 80GB 1 fp16 16 80 $1.50
NVIDIA H100 PCIe 80GB 1 fp8 8 80 $1.50
NVIDIA H100 SXM 80GB 1 fp16 16 80 $1.90
NVIDIA H100 SXM 80GB 1 fp8 8 80 $1.90
NVIDIA H200 141GB 1 fp16 16 141 $3.20
NVIDIA H200 141GB 1 fp8 8 141 $3.20
NVIDIA B200 192GB 1 fp16 16 192 $5.00
NVIDIA B200 192GB 1 fp8 8 192 $5.00

Llama 3.1 8B Instruct · 8.0B · 131,072 ctx

GPU×QuantWeights GBVRAM GB$/hr
NVIDIA GeForce RTX 3090 1 fp16 16 24 $0.20
NVIDIA GeForce RTX 3090 1 fp8 8 24 $0.20
NVIDIA GeForce RTX 4090 1 fp16 16 24 $0.35
NVIDIA GeForce RTX 4090 1 fp8 8 24 $0.35
NVIDIA RTX A6000 1 fp16 16 48 $0.45
NVIDIA RTX A6000 1 fp8 8 48 $0.45
NVIDIA L40S 1 fp16 16 48 $0.55
NVIDIA L40S 1 fp8 8 48 $0.55
NVIDIA A100 80GB 1 fp16 16 80 $1.27
NVIDIA A100 80GB 1 fp8 8 80 $1.27
NVIDIA H100 PCIe 80GB 1 fp16 16 80 $1.50
NVIDIA H100 PCIe 80GB 1 fp8 8 80 $1.50
NVIDIA H100 SXM 80GB 1 fp16 16 80 $1.90
NVIDIA H100 SXM 80GB 1 fp8 8 80 $1.90
NVIDIA H200 141GB 1 fp16 16 141 $3.20
NVIDIA H200 141GB 1 fp8 8 141 $3.20
NVIDIA B200 192GB 1 fp16 16 192 $5.00
NVIDIA B200 192GB 1 fp8 8 192 $5.00

Mistral Small 3.2 24B Instruct · 24.0B · 128,000 ctx

GPU×QuantWeights GBVRAM GB$/hr
NVIDIA GeForce RTX 3090 2 fp8 24 48 $0.40
NVIDIA RTX A6000 1 fp8 24 48 $0.45
NVIDIA L40S 1 fp8 24 48 $0.55
NVIDIA GeForce RTX 4090 2 fp8 24 48 $0.70
NVIDIA GeForce RTX 3090 4 fp16 48 96 $0.80
NVIDIA RTX A6000 2 fp16 48 96 $0.90
NVIDIA L40S 2 fp16 48 96 $1.10
NVIDIA A100 80GB 1 fp8 24 80 $1.27
NVIDIA A100 80GB 1 fp16 48 80 $1.27
NVIDIA GeForce RTX 4090 4 fp16 48 96 $1.40
NVIDIA H100 PCIe 80GB 1 fp16 48 80 $1.50
NVIDIA H100 PCIe 80GB 1 fp8 24 80 $1.50
NVIDIA H100 SXM 80GB 1 fp16 48 80 $1.90
NVIDIA H100 SXM 80GB 1 fp8 24 80 $1.90
NVIDIA H200 141GB 1 fp16 48 141 $3.20
NVIDIA H200 141GB 1 fp8 24 141 $3.20
NVIDIA B200 192GB 1 fp16 48 192 $5.00
NVIDIA B200 192GB 1 fp8 24 192 $5.00

Gemma 3 27B IT · 27.4B · 131,072 ctx

GPU×QuantWeights GBVRAM GB$/hr
NVIDIA GeForce RTX 3090 2 fp8 28 48 $0.40
NVIDIA RTX A6000 1 fp8 28 48 $0.45
NVIDIA L40S 1 fp8 28 48 $0.55
NVIDIA GeForce RTX 4090 2 fp8 28 48 $0.70
NVIDIA GeForce RTX 3090 4 fp16 55 96 $0.80
NVIDIA RTX A6000 2 fp16 55 96 $0.90
NVIDIA L40S 2 fp16 55 96 $1.10
NVIDIA A100 80GB 1 fp8 28 80 $1.27
NVIDIA A100 80GB 1 fp16 55 80 $1.27
NVIDIA GeForce RTX 4090 4 fp16 55 96 $1.40
NVIDIA H100 PCIe 80GB 1 fp16 55 80 $1.50
NVIDIA H100 PCIe 80GB 1 fp8 28 80 $1.50
NVIDIA H100 SXM 80GB 1 fp16 55 80 $1.90
NVIDIA H100 SXM 80GB 1 fp8 28 80 $1.90
NVIDIA H200 141GB 1 fp16 55 141 $3.20
NVIDIA H200 141GB 1 fp8 28 141 $3.20
NVIDIA B200 192GB 1 fp16 55 192 $5.00
NVIDIA B200 192GB 1 fp8 28 192 $5.00

DeepSeek R1 Distill Qwen 32B · 32.8B · 131,072 ctx

GPU×QuantWeights GBVRAM GB$/hr
NVIDIA GeForce RTX 3090 2 fp8 33 48 $0.40
NVIDIA RTX A6000 1 fp8 33 48 $0.45
NVIDIA L40S 1 fp8 33 48 $0.55
NVIDIA GeForce RTX 4090 2 fp8 33 48 $0.70
NVIDIA GeForce RTX 3090 4 fp16 66 96 $0.80
NVIDIA RTX A6000 2 fp16 66 96 $0.90
NVIDIA L40S 2 fp16 66 96 $1.10
NVIDIA A100 80GB 1 fp8 33 80 $1.27
NVIDIA A100 80GB 1 fp16 66 80 $1.27
NVIDIA GeForce RTX 4090 4 fp16 66 96 $1.40
NVIDIA H100 PCIe 80GB 1 fp16 66 80 $1.50
NVIDIA H100 PCIe 80GB 1 fp8 33 80 $1.50
NVIDIA H100 SXM 80GB 1 fp16 66 80 $1.90
NVIDIA H100 SXM 80GB 1 fp8 33 80 $1.90
NVIDIA H200 141GB 1 fp16 66 141 $3.20
NVIDIA H200 141GB 1 fp8 33 141 $3.20
NVIDIA B200 192GB 1 fp16 66 192 $5.00
NVIDIA B200 192GB 1 fp8 33 192 $5.00

Qwen3 30B-A3B Instruct 2507 · 30.5B · 262,144 ctx

GPU×QuantWeights GBVRAM GB$/hr
NVIDIA GeForce RTX 3090 2 fp8 31 48 $0.40
NVIDIA RTX A6000 1 fp8 31 48 $0.45
NVIDIA L40S 1 fp8 31 48 $0.55
NVIDIA GeForce RTX 4090 2 fp8 31 48 $0.70
NVIDIA GeForce RTX 3090 4 fp16 61 96 $0.80
NVIDIA RTX A6000 2 fp16 61 96 $0.90
NVIDIA L40S 2 fp16 61 96 $1.10
NVIDIA A100 80GB 1 fp8 31 80 $1.27
NVIDIA A100 80GB 1 fp16 61 80 $1.27
NVIDIA GeForce RTX 4090 4 fp16 61 96 $1.40
NVIDIA H100 PCIe 80GB 1 fp16 61 80 $1.50
NVIDIA H100 PCIe 80GB 1 fp8 31 80 $1.50
NVIDIA H100 SXM 80GB 1 fp16 61 80 $1.90
NVIDIA H100 SXM 80GB 1 fp8 31 80 $1.90
NVIDIA H200 141GB 1 fp16 61 141 $3.20
NVIDIA H200 141GB 1 fp8 31 141 $3.20
NVIDIA B200 192GB 1 fp16 61 192 $5.00
NVIDIA B200 192GB 1 fp8 31 192 $5.00

gpt-oss 120B · 117.0B · 131,072 ctx

GPU×QuantWeights GBVRAM GB$/hr
NVIDIA GeForce RTX 3090 8 fp8 117 192 $1.60
NVIDIA RTX A6000 4 fp8 117 192 $1.80
NVIDIA L40S 4 fp8 117 192 $2.20
NVIDIA A100 80GB 2 fp8 117 160 $2.54
NVIDIA GeForce RTX 4090 8 fp8 117 192 $2.80
NVIDIA H100 PCIe 80GB 2 fp8 117 160 $3.00
NVIDIA H200 141GB 1 fp8 117 141 $3.20
NVIDIA RTX A6000 8 fp16 234 384 $3.60
NVIDIA H100 SXM 80GB 2 fp8 117 160 $3.80
NVIDIA L40S 8 fp16 234 384 $4.40
NVIDIA B200 192GB 1 fp8 117 192 $5.00
NVIDIA A100 80GB 4 fp16 234 320 $5.08
NVIDIA H100 PCIe 80GB 4 fp16 234 320 $6.00
NVIDIA H200 141GB 2 fp16 234 282 $6.40
NVIDIA H100 SXM 80GB 4 fp16 234 320 $7.60
NVIDIA B200 192GB 2 fp16 234 384 $10.00

Qwen2.5 72B Instruct · 72.7B · 131,072 ctx

GPU×QuantWeights GBVRAM GB$/hr
NVIDIA GeForce RTX 3090 4 fp8 73 96 $0.80
NVIDIA RTX A6000 2 fp8 73 96 $0.90
NVIDIA L40S 2 fp8 73 96 $1.10
NVIDIA GeForce RTX 4090 4 fp8 73 96 $1.40
NVIDIA GeForce RTX 3090 8 fp16 146 192 $1.60
NVIDIA RTX A6000 4 fp16 146 192 $1.80
NVIDIA L40S 4 fp16 146 192 $2.20
NVIDIA A100 80GB 2 fp8 73 160 $2.54
NVIDIA GeForce RTX 4090 8 fp16 146 192 $2.80
NVIDIA H100 PCIe 80GB 2 fp8 73 160 $3.00
NVIDIA H200 141GB 1 fp8 73 141 $3.20
NVIDIA H100 SXM 80GB 2 fp8 73 160 $3.80
NVIDIA B200 192GB 1 fp16 146 192 $5.00
NVIDIA B200 192GB 1 fp8 73 192 $5.00
NVIDIA A100 80GB 4 fp16 146 320 $5.08
NVIDIA H100 PCIe 80GB 4 fp16 146 320 $6.00
NVIDIA H200 141GB 2 fp16 146 282 $6.40
NVIDIA H100 SXM 80GB 4 fp16 146 320 $7.60

Llama 3.3 70B Instruct · 70.6B · 131,072 ctx

GPU×QuantWeights GBVRAM GB$/hr
NVIDIA GeForce RTX 3090 4 fp8 71 96 $0.80
NVIDIA RTX A6000 2 fp8 71 96 $0.90
NVIDIA L40S 2 fp8 71 96 $1.10
NVIDIA GeForce RTX 4090 4 fp8 71 96 $1.40
NVIDIA GeForce RTX 3090 8 fp16 142 192 $1.60
NVIDIA RTX A6000 4 fp16 142 192 $1.80
NVIDIA L40S 4 fp16 142 192 $2.20
NVIDIA A100 80GB 2 fp8 71 160 $2.54
NVIDIA GeForce RTX 4090 8 fp16 142 192 $2.80
NVIDIA H100 PCIe 80GB 2 fp8 71 160 $3.00
NVIDIA H200 141GB 1 fp8 71 141 $3.20
NVIDIA H100 SXM 80GB 2 fp8 71 160 $3.80
NVIDIA B200 192GB 1 fp16 142 192 $5.00
NVIDIA B200 192GB 1 fp8 71 192 $5.00
NVIDIA A100 80GB 4 fp16 142 320 $5.08
NVIDIA H100 PCIe 80GB 4 fp16 142 320 $6.00
NVIDIA H200 141GB 2 fp16 142 282 $6.40
NVIDIA H100 SXM 80GB 4 fp16 142 320 $7.60

Qwen3 235B-A22B Instruct 2507 · 235.0B · 262,144 ctx

GPU×QuantWeights GBVRAM GB$/hr
NVIDIA RTX A6000 8 fp8 235 384 $3.60
NVIDIA L40S 8 fp8 235 384 $4.40
NVIDIA A100 80GB 4 fp8 235 320 $5.08
NVIDIA H100 PCIe 80GB 4 fp8 235 320 $6.00
NVIDIA H200 141GB 2 fp8 235 282 $6.40
NVIDIA H100 SXM 80GB 4 fp8 235 320 $7.60
NVIDIA B200 192GB 2 fp8 235 384 $10.00
NVIDIA A100 80GB 8 fp16 470 640 $10.16
NVIDIA H100 PCIe 80GB 8 fp16 470 640 $12.00
NVIDIA H200 141GB 4 fp16 470 564 $12.80
NVIDIA H100 SXM 80GB 8 fp16 470 640 $15.20
NVIDIA B200 192GB 4 fp16 470 768 $20.00

Mixtral 8x7B Instruct v0.1 · 46.7B · 32,768 ctx

GPU×QuantWeights GBVRAM GB$/hr
NVIDIA GeForce RTX 3090 4 fp8 47 96 $0.80
NVIDIA RTX A6000 2 fp8 47 96 $0.90
NVIDIA L40S 2 fp8 47 96 $1.10
NVIDIA A100 80GB 1 fp8 47 80 $1.27
NVIDIA GeForce RTX 4090 4 fp8 47 96 $1.40
NVIDIA H100 PCIe 80GB 1 fp8 47 80 $1.50
NVIDIA GeForce RTX 3090 8 fp16 94 192 $1.60
NVIDIA RTX A6000 4 fp16 94 192 $1.80
NVIDIA H100 SXM 80GB 1 fp8 47 80 $1.90
NVIDIA L40S 4 fp16 94 192 $2.20
NVIDIA A100 80GB 2 fp16 94 160 $2.54
NVIDIA GeForce RTX 4090 8 fp16 94 192 $2.80
NVIDIA H100 PCIe 80GB 2 fp16 94 160 $3.00
NVIDIA H200 141GB 1 fp16 94 141 $3.20
NVIDIA H200 141GB 1 fp8 47 141 $3.20
NVIDIA H100 SXM 80GB 2 fp16 94 160 $3.80
NVIDIA B200 192GB 1 fp16 94 192 $5.00
NVIDIA B200 192GB 1 fp8 47 192 $5.00