TOP_UP
Pay ¥1 to receive $1 in platform credit. API usage is deducted directly using each channel's published pricing factor.
PRICE_TABLE
Long-context rates apply to the entire request when input plus cached input exceeds 272K tokens.
gpt-6-astra
The flagship GPT-6 Astra model for computer use, browsing, and the hardest software engineering work. 1.05M context, billed at one rate across the whole window with no long-context tier.
This model is available on the GPT Pro (0.22x) and Anti-Downgrade channels. The GPT Economy channel does not carry it; smart keys are routed to GPT Pro and billed at its rates.
gpt-6.1-sol
GPT-6.1 Sol balances capability and price for coding and professional work. 1.05M context; long-context rates apply above 272K input tokens.
gpt-6-sol
GPT-6 Sol balances capability and price for coding and professional work. 1.05M context; long-context rates apply above 272K input tokens.
gpt-6-luna
The fastest and most cost-efficient GPT-6 model for everyday agent tasks. 1.05M context.
This model is available on the Luna channel (0.7x) only. The GPT Pro, GPT Economy, Anti-Downgrade, smart, and unrestricted channels do not carry it; create a Luna channel API key to use it.
