Skip to main content
In Cascade, you can easily switch between different models of your choosing. Depending on the model you select, each of your input prompts will consume a different number of prompt credits.
Model prompt credits are not applicable in Cognition Platform plans. For Cognition Platform customers, please refer to the ACU consumption shown in your Cognition Platform settings.
Under the text input box, you will see a model selection dropdown menu containing the following models:
For the most up-to-date pricing and availability, please refer to the model selector in Cascade within the Windsurf IDE.
ModelCreditsFreeProTeamsEnterpriseTrial
SWE-1.5 iconSWE-1.5 iconSWE-1.5
0
Gift iconGift icon
AvailableAvailableAvailableAvailableAvailableAvailableAvailableAvailableAvailableAvailable
Claude Opus 4.6 iconClaude Opus 4.6 iconClaude Opus 4.6
6
AvailableAvailableAvailableAvailableAvailableAvailableAvailableAvailable
Claude Sonnet 4.6 iconClaude Sonnet 4.6 iconClaude Sonnet 4.6
4
AvailableAvailableAvailableAvailableAvailableAvailableAvailableAvailable
Claude Opus 4.5 iconClaude Opus 4.5 iconClaude Opus 4.5
4
AvailableAvailableAvailableAvailable6AvailableAvailable
Claude Sonnet 4.5 iconClaude Sonnet 4.5 iconClaude Sonnet 4.5
2
Gift iconGift icon
AvailableAvailableAvailableAvailable3AvailableAvailable
GPT-5.2-Codex (Low Reasoning Fast) iconGPT-5.2-Codex (Low Reasoning Fast) iconGPT-5.2-Codex (Low Reasoning Fast)
2
AvailableAvailableAvailableAvailableAvailableAvailableAvailableAvailable
GPT-5.3-Codex (Medium Reasoning) iconGPT-5.3-Codex (Medium Reasoning) iconGPT-5.3-Codex (Medium Reasoning)
2
AvailableAvailableAvailableAvailableAvailableAvailableAvailableAvailable

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 frontier model, SWE-1.5, achieves near-SOTA performance in a fraction of the time. Our in house models include:
  • SWE-1.5: Our best agentic coding model we’ve released. Near Claude 4.5-level performance, at 13x the speed. Read our research announcement.
  • 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

Bring your own key (BYOK)

This is only available to free and paid individual users.
For certain models, we allow users to bring their own API keys. In the model dropdown menu, individual users will see models labled with BYOK. To input your API key, navigate to this page in the subscription settings and add your key. If you have not configured your API key, it will return an error if you try to use the BYOK model. Currently, we only support BYOK for these models:
  • Claude 4 Sonnet
  • Claude 4 Sonnet (Thinking)
  • Claude 4 Opus
  • Claude 4 Opus (Thinking)
Gift icon
Available
Available
Available
Available
Available
Claude Opus 4.6 icon
Available
Available
Available
Available
Claude Sonnet 4.6 icon
Available
Available
Available
Available
Claude Opus 4.5 icon
Available
Available
Available
Claude Sonnet 4.5 icon
Gift icon
Available
Available
Available
GPT-5.2-Codex (Low Reasoning Fast) icon
Available
Available
Available
Available
GPT-5.3-Codex (Medium Reasoning) icon
Available
Available
Available
Available
SWE-1.5 icon