CrofAI - Pricing

Model

Quantization Context / Max output

$ / M in

$ / M out

Speed

kimi-k2-0905

fp8 131,072 / 131,072

$0.15

$0.55

~29 t/s

kimi-k2-0905-turbo

fp8 131,072 / 8,192

$0.35

$1.00

~1019 t/s

kimi-k2-eco

fp4 131,072 / 131,072

$0.05

$0.10

~15 t/s

glm-4.6

fp8 131,072 / 131,072

$0.30

$0.60

~51 t/s

glm-4.5

fp8 131,072 / 131,072

$0.20

$0.40

~25 t/s

deepseek-v3.2-exp

fp4 131,072 / 131,072

$0.15

$0.30

~107 t/s

deepseek-v3.1-terminus

fp4 131,072 / 131,072

$0.20

$0.50

~18 t/s

deepseek-v3.1-terminus-reasoner

fp4 131,072 / 131,072

$0.20

$0.50

~36 t/s

deepseek-v3.1

fp4 131,072 / 131,072

$0.15

$0.50

~18 t/s

deepseek-v3.1-reasoner

fp4 131,072 / 131,072

$0.15

$0.50

~42 t/s

deepseek-v3-0324

fp4 131,072 / 8,192

$0.20

$0.25

~14 t/s

deepseek-v3-0324-turbo

fp4 131,072 / 8,192

$0.50

$1.00

~839 t/s

deepseek-r1-0528

fp4 131,072 / 131,072

$0.25

~209 t/s

deepseek-r1-0528-turbo

fp4 131,072 / 131,072

$1.00

$2.00

~819 t/s

qwen3-next-80b-a3b-instruct

fp8 262,144 / 262,144

$0.08

$0.38

~282 t/s

qwen3-235b-a22b-2507-instruct

fp8 131,072 / 131,072

$0.10

$0.25

~20 t/s

qwen3-235b-a22b-2507-thinking

fp8 131,072 / 131,072

$0.10

$0.30

~29 t/s

qwen3-coder

fp8 131,072 / 131,072

$0.15

$0.35

~256 t/s

qwen3-coder-turbo

fp8 131,072 / 131,072

$0.20

$0.50

~439 t/s

gpt-oss-120b

fp4 131,072 / 131,072

$0.07

$0.27

~132 t/s

gemma-3-27b-it

fp8 131,072 / 131,072

$0.04

$0.10

~246 t/s

llama-4-scout

fp8 262,144 / 16,384

$0.08

$0.40

~39 t/s

llama3.3-70b

fp8 131,072 / 8,192

$0.12

$0.20

~68 t/s

deepseek-r1-distill-llama-70b

fp4 65,536 / 65,536

$0.10

~62 t/s

deepseek-r1-distill-qwen-32b

fp4 65,536 / 65,536

$0.10

~40 t/s