Supported Models
The following models are currently available in the Brivionix model marketplace. All models can be called through the OpenAI-compatible endpoint:
POST https://brivionix.com/v1/chat/completionsModel Overview
| Model | Context | Capabilities | Input Price | Output Price | Cache Read |
|---|---|---|---|---|---|
gpt-5.2 | 400K | Reasoning, Tools, Files, Vision | $2.0590 | $16.4700 | $0.2060 |
gpt-5.3-codex | 400K | Reasoning, Tools, Vision | $2.0590 | $16.4700 | $0.2060 |
gpt-5.4 | 400K | Reasoning, Tools, Files, Vision | $2.9410 | $17.6470 | $0.2940 |
gpt-5.4-mini | 400K | Reasoning, Tools, Files, Vision | $2.3410 | $14.0470 | $0.2140 |
gpt-5.5 | 1.1M | Reasoning, Tools, Files, Vision | $5.8820 | $47.0590 | $0.5880 |
Prices are shown in USD / 1M Tokens, matching the default model marketplace display.
Actual Cost
Enable Top Up price in the model marketplace to view the estimated real billing cost based on the top-up rate 0.85 RMB = 1 API Credit.
Capability Notes
| Label | Meaning |
|---|---|
| Reasoning | Advanced reasoning |
| Tools | Tool calling / Function Calling |
| Files | File-related capability |
| Vision | Image understanding / multimodal input |
| 400K / 1.1M | Context length |
Groups
The public model list is documented for the default group. When creating or editing a token, use the default group.
Actual access also depends on the token's Available models setting:
- Empty: no extra model whitelist; the token can call available models in the
defaultgroup. - Filled: only the listed models can be called, and the model name must exactly match the table above.
If the API returns "model unavailable" or "permission denied", first check that the token group is default, the model name is correct, and Available models includes the target model.
Request Example
curl https://brivionix.com/v1/chat/completions \
-H "Authorization: Bearer sk-your-api-key" \
-H "Content-Type: application/json" \
-d '{
"model": "gpt-5.4-mini",
"messages": [{"role": "user", "content": "Hello"}]
}'See Chat Completions for more request parameters.
