# Sources (accessed 2026-09-28):
# https://www.alibabacloud.com/help/en/model-studio/model-pricing (Singapore)
# https://www.alibabacloud.com/help/en/model-studio/context-cache
# https://www.qwencloud.com/models/qwen3.5-flash
# Toggle: enable_thinking true|false
# Budget: thinking_budget (integer reasoning tokens, maximum 81920)
base_model = "alibaba/qwen3.5-flash"
reasoning_options = [{ type = "toggle" }, { type = "budget_tokens", max = 81_920 }]

[interleaved]
field = "reasoning_content"

[cost]
input = 0.10
output = 0.40
cache_read = 0.01
cache_write = 0.125