> ## Documentation Index
> Fetch the complete documentation index at: https://docs.devinenterprise.com/llms.txt
> Use this file to discover all available pages before exploring further.

# AI Models

> Available AI models in Devin Desktop Cascade including SWE-1.6, SWE-1.5, Claude, and GPT. Compare model capabilities, credit costs, and performance.

<Card title="Adaptive" icon="shuffle" href="/desktop/adaptive" horizontal={true}>
  For most users, we recommend **Adaptive** — our intelligent model router that automatically selects the best model for each task, delivering the right level of intelligence for every prompt.
</Card>

In Cascade, you can easily switch between different models of your choosing.

Under the text input box, you will see a model selection dropdown menu containing the following models:

<Info>For the most up-to-date pricing and availability, please refer to the model selector in Cascade within the Windsurf IDE.</Info>

<Tabs>
  <Tab title="Self-serve">
    Your quota and extra usage is billed based on the token cost of the model you select. You can view the cost of each model in the table below.
  </Tab>

  <Tab title="Enterprise (ACUs)">
    Model usage is converted to ACUs based on the per-token rates below.
  </Tab>

  <Tab title="Legacy enterprise (credits)">
    This only applies to credit-based enterprise customers. Newer enterprise plans are billed in ACUs — see the Enterprise (ACUs) tab.
  </Tab>
</Tabs>

# SWE-1.6, SWE-1.5, swe-grep, SWE-1

Our SWE model family of in-house frontier models are built specifically for software engineering tasks.

Our latest model, SWE-1.6, is generally available in Windsurf and is optimized for both intelligence and model UX. SWE-1.6 Fast is industry-leading in speed.

Our in-house models include:

* `SWE-1.6`: Our latest model built for software engineering agents, optimized for both intelligence and model UX. Achieves comparable SWE-Bench Pro performance to the SWE-1.6 Preview, which improved on SWE-1.5 by more than 10%. Uses parallel tool calls more often, loops far less, and relies more on its own tools than the terminal, leading to more efficient trajectories and a smoother user experience. Read our [research announcement](https://cognition.ai/blog/swe-1-6).
* `SWE-1.6 Fast`: A faster version of SWE-1.6 available to paying users, delivering the same intelligence with unmatched speed and cost.
* `SWE-1.5`: Our previous frontier agentic coding model. Near Claude 4.5-level performance at 13x the speed. Read our [research announcement](https://cognition.ai/blog/swe-1-5).
* `SWE-1`: Our first agentic coding model. Achieved Claude 3.5-level performance at a fraction of the cost.
* `SWE-1-mini`: Powers passive suggestions in Windsurf Tab, optimized for real-time latency.
* `swe-grep`: Powers context retrieval and [Fast Context](/desktop/context-awareness/fast-context)