> For the complete documentation index, see [llms.txt](https://references.everstrike.io/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://references.everstrike.io/ai/models.md).

# Models

Choose a model for each AI strategy. Model choice affects evaluation speed and credit cost.

### Available models

| Model                        | Typical speed | Cost per evaluation | Description                              |
| ---------------------------- | ------------: | ------------------: | ---------------------------------------- |
| GPT-4o-mini                  |            1s |           $0.000125 | Oldest model.                            |
| GPT-4.1                      |          1.5s |             $0.0015 | Fast, expensive, and powerful.           |
| GPT-4.1-mini                 |            1s |            $0.00015 | Default model. Best for most strategies. |
| GPT-4.1-nano                 |          0.5s |           $0.000075 | Very fast and inexpensive.               |
| GPT-5-mini-minimal-reasoning |            3s |             $0.0002 | Balanced.                                |
| GPT-5-nano-minimal-reasoning |            3s |            $0.00005 | Inexpensive.                             |
| GPT-5-mini-low-reasoning     |            5s |             $0.0004 | Moderately powerful.                     |
| GPT-5-nano-low-reasoning     |            5s |             $0.0001 | Inexpensive and moderately powerful.     |
| GPT-5-mini-medium-reasoning  |           10s |             $0.0008 | Powerful.                                |
| GPT-5-nano-medium-reasoning  |           10s |             $0.0002 | Inexpensive and powerful.                |
| GPT-5-mini-high-reasoning    |           20s |              $0.002 | Very powerful and expensive.             |
| GPT-5-nano-high-reasoning    |           20s |             $0.0004 | Powerful and inexpensive.                |

### Evaluation frequency limit

Strategies with an evaluation frequency of 15 seconds or less can use only `GPT-4.1-nano`.

### Speed and cost notes

Listed speeds are averages, not guarantees. Evaluation time also depends on the prompt and the data available to the model.

Restricting strategy data can reduce the cost per evaluation by up to 50%. See [Supported Data](/ai/supported-data.md) to choose available data points.

See [Strategies](/ai/strategies.md) for model configuration and strategy cost estimates.