GPT-5.4 API pricing
A well-rounded mainline GPT-5.x model for general-purpose workloads.
Pricing details
The table shows the lowest group-discounted price in USD per 1 million tokens. It uses the lowest multiplier among this model’s enabled groups: gpt纯血pro at 0.2×. Available groups: gpt混池 / gpt纯血pro. If your account belongs to a different group, your charge will differ. Always use the current price shown in the console as the billing authority.
Scroll horizontally to view the full table.
| Item | Price (lowest group) | Description |
|---|---|---|
| Input (prompt) | $0.5 / 1M tokens and up | Includes the gpt纯血pro group multiplier |
| Output (completion) | $3 / 1M tokens and up | Includes the gpt纯血pro group multiplier |
| Billing group | gpt纯血pro | 0.2×, the lowest multiplier available for this model |
| Context window | 256,000 tokens | Common published specification; the model’s actual upstream limit applies. |
| Model ID | gpt-5.4 | |
API example
Call GPT-5.4 with the OpenAI-compatible API format:
curl https://qionggeme.com/v1/chat/completions \
-H "Content-Type: application/json" \
-H "Authorization: Bearer $YOUR_API_KEY" \
-d '{
"model": "gpt-5.4",
"messages": [
{"role": "user", "content": "Hello"}
]
}'Frequently asked questions
How much does the GPT-5.4 API cost?
GPT-5.4 (gpt-5.4) is displayed at the lowest enabled group rate: about $0.5 per 1 million input tokens and $3 per 1 million output tokens through group gpt纯血pro at 0.2×. Your actual charge depends on your account’s current group; check the console for the billing price.
What context window does GPT-5.4 support?
GPT-5.4 commonly supports a context window of about 256,000 tokens. The actual upstream model limit applies.
How do I connect to the GPT-5.4 API?
Create an API key in qionggeme, set the API Base to https://qionggeme.com—usually https://qionggeme.com/v1 in OpenAI-compatible clients—and use model ID gpt-5.4. Do not use the content-site URL https://qionggeme.com/docs.