Agentic Trust supports multiple AI models from leading providers. You can select a model per project from the Settings page in the dashboard. The default model is Claude Sonnet 4.6.
Model pricing
All prices are per million tokens. A dash (—) means the provider does not charge for that tier.
| Name | Input | Cache Write | Cache Read | Output |
|---|
| Sonnet 4.6 | $3 | $3.75 | $0.30 | $15 |
| Haiku 4.5 | $1 | $1.25 | $0.10 | $5 |
| GPT-5.4 | $2.50 | — | $0.25 | $15 |
| GPT-5.2 | $1.75 | — | $0.17 | $14 |
| Gemini 3.1 Flash Lite | $0.25 | $0.08 | $0.02 | $1.50 |
| GLM 5 | $0.60 | — | $0.12 | $1.90 |
| MiniMax M2.5 | $0.29 | — | $0.03 | $1.20 |
| Kimi K2.5 | $0.41 | — | $0.07 | $2.06 |
Choosing a model
Start with Sonnet 4.6 (the default). Switch to Haiku 4.5 if you need faster responses at lower cost, or GPT-5.4 for maximum capability.
| Use case | Recommended model | Why |
|---|
| General customer support | Sonnet 4.6 | Best all-around quality |
| High-volume, simple Q&A | Haiku 4.5 | Fastest, lowest cost |
| Complex reasoning tasks | GPT-5.4 | Adaptive reasoning with configurable effort |
| Large context / long docs | GPT-5.4 | 1.05M context window |
| Code generation | GPT-5.2 or GPT-5.4 | Strong at coding and analysis |
| Budget-conscious deployment | Gemini 3.1 Flash Lite | Cheapest input/output |
| Long document generation | GLM 5 or GPT-5.2 | Highest max output tokens |
Changing the model
Go to Settings in your project dashboard and select a model from the Model dropdown. The change takes effect for new conversations immediately. 
