base_model = "deepseek/deepseek-r1"
reasoning_options = []
structured_output = true

[cost]
input = 0.7
output = 2.5

[limit]
context = 64_000