- Model catalog
Model catalog
One credit unit across every model. Input settles at the listed multiplier, while cached input and output use model-specific weights. Rates are read live from the gateway rate card.
- Models
- 18
- Context
- Varies
- Billing
- Per token
| Model | Context (tokens) | Input | Cache | Output | Status |
|---|---|---|---|---|---|
claude-opus-4.7vision | 200K | 10 | 5 | 50 | Operational |
deepseek-v4-mod | 0 | 6 | 1.8 | 30 | Operational |
deepseek-v4-pro | 1M | 3 | 1.5 | 15 | Operational |
deepseek-v4.1-flash | 1M | 0.75 | 0.375 | 3.75 | Operational |
deepseek-v4.1-flash-1017-modvision | 0 | 6 | 3 | 30 | Operational |
glm-5.2 | 200K | 2.5 | 1.25 | 10 | Operational |
glm-5.2-mod | 0 | 6 | 3 | 24 | Operational |
glm-5.3 | 200K | 2.5 | 0.75 | 10 | Operational |
glm-5.3-flash | 200K | 2 | 1 | 10 | Operational |
glm-5.3-flash-modvision | 0 | 6 | 3 | 30 | Operational |
glm-5.3-flashx-modvision | 0 | 6 | 3 | 30 | Operational |
glm-5.3-mod | 0 | 6 | 3 | 30 | Operational |
image-gemini-2.5-flash-imagevision | 0 | 2 | 1 | 10 | Operational |
image-gemini-3.0-pro-imagevision | 0 | 2 | 1 | 10 | Operational |
image-gemini-3.1-flash-imagevision | 0 | 2 | 1 | 10 | Operational |
image-gpt-image-2vision | 0 | 2 | 1 | 10 | Operational |
kimi-k3vision | 1.0M | 4 | 2 | 16 | Operational |
kimi-k3-mod | 0 | 6 | 3 | 24 | Operational |
Rates are live credits per token from the current rate card. Cached and output weights are model-specific when configured. See the billing docs for the full formula.