How to Use GPT-4.1 in Windsurf
GPT-4.1 is useful for code edits, detailed instructions, and analysis of long documents. Its 1,047,576-token context window is much larger than its 32,768-token output limit, so ask for focused results.
Prompts to try with GPT-4.1 in Windsurf
Copy an example into Windsurf, or ask in your own words.
Use GPT-4.1 to investigate the failing tests in this project. Read the relevant implementation and test output, fix the underlying bug, and run the affected tests. Report the changed files and verification results.
Use GPT-4.1 to review the current branch diff with its surrounding code. Report only actionable bugs introduced by the changes, with file locations and a concrete failure scenario. Do not edit anything.
Use GPT-4.1 to compare the README instructions with the implementation. Identify commands or examples that no longer match the code, then propose a unified diff limited to documentation.
Use GPT-4.1 to trace request validation through this repository. Identify where malformed input is accepted, propose a focused patch, and add a regression test for the demonstrated failure.
Use GPT-4.1 to refactor the longest duplicated validation logic in this project. Preserve public behavior, follow the existing conventions, and return a unified diff with the relevant test results.
How to set up GPT-4.1 in Windsurf
Pick the way that suits you. Each takes about two minutes.
Use Spiral's MCP server
- Add the server
Open the Customize page from the Agents window and click MCPs. Under This computer, open the config file, or edit it directly: ~/.config/devin/mcp_config.json on macOS and Linux, %APPDATA%\devin\mcp_config.json on Windows. Add this:
{ "mcpServers": { "spiral": { "url": "https://api.spiral.ai/mcp" } } } - Sign in
Click Authenticate on the Spiral server in Customize › MCPs, or let the agent call Spiral once: your browser opens. Sign in to Spiral, pick the organization and click Authorize. No API key.
devin mcp login spiral
- Check it
Spiral is listed under This computer in Customize › MCPs. From the terminal:
devin mcp list
Use the Spiral skill
- Send Windsurf this message
Your agent reads Spiral's skill and sets itself up.
set up https://spiral.ai/SKILL.md
- Sign in or paste a key
Approve Spiral in the browser, or give the agent a key from the console.
Use the Spiral CLI
- Install and sign in
Installs the Spiral CLI, signs you in and adds its skill to your coding agents.
npx -y spiral-cli@latest init
Call the Spiral API
- Create an API key
In the console, under API keys.
- Run GPT-4.1
One request; the result comes back by polling or a webhook.
curl https://api.spiral.ai/v1/runs \ -H "Authorization: Bearer $SPIRAL_API_KEY" \ -H "Content-Type: application/json" \ -H "Idempotency-Key: $(uuidgen)" \ -d '{ "provider": "openai", "model": "openai/gpt-4.1", "input": { "messages": [ { "role": "system" } ] } }'
Then paste one of the prompts above: Windsurf finds GPT-4.1 on Spiral, shows the price and runs it.
GPT-4.1 providers and prices
Spiral runs GPT-4.1 through OpenAI. Open a card for today's price, with every fee included.
Up to $0.04269 per run at the default inputs. $2.60 per 1M input tokens, $0.65 per 1M cached input tokens, $10.40 per 1M output tokens.
Settings you can ask for
Ask Windsurf for any of these in your prompt.
| Setting | Allowed values | Default | Description |
|---|---|---|---|
| Seed | a whole number | ||
| Top p | 0 to 1 | ||
| Logprobs | on or off | ||
| Max tokens | a whole number from 16 | The most output tokens to generate. | |
| Temperature | 0 to 2 | ||
| Top logprobs | 0 to 20 | ||
| Presence penalty | -2 to 2 | ||
| Frequency penalty | -2 to 2 | ||
| Parallel tool calls | on or off |
Every API field of GPT-4.1 is on its tool page: open its card above.
What can you do with GPT-4.1 in Windsurf?
GPT-4.1 can turn repository context into focused patches and code review findings. Give it explicit boundaries and a required diff format to guide the changes.
Focused bug fixes
Ask GPT-4.1 to trace a reported failure, change the relevant code, and return a patch with verification results.
Pull request review
Give GPT-4.1 a diff and surrounding code. Request actionable findings with a concrete failure scenario for each.
Questions across long documents
Use GPT-4.1 to locate relevant passages across supplied material and answer a narrow question with supporting excerpts.
Structured record extraction
Have GPT-4.1 turn unstructured text into records, leaving unsupported values empty and retaining evidence for review.
Policy-based response drafts
Ask GPT-4.1 to draft replies using supplied policies, explicit exceptions, and a defined response format.
Recurring source digests
Use GPT-4.1 to select relevant updates from supplied text and produce a digest with source references.
Search result ranking
Give GPT-4.1 a question and candidate passages. Request a ranked shortlist with a relevance explanation for each selection.
Constrained text rewrites
Ask GPT-4.1 to rewrite supplied text while preserving specified facts, following a length target, and avoiding named terms.
More on GPT-4.1 and Windsurf
Overview
GPT-4.1 in other agents
Questions about GPT-4.1 in Windsurf
Can Windsurf use GPT-4.1?
Yes. Add Spiral to Windsurf as an MCP server, and Windsurf can find and run GPT-4.1 like any built-in tool.
How much does GPT-4.1 cost in Windsurf?
Up to $0.04269 per run at the default inputs. $2.60 per 1M input tokens, $0.65 per 1M cached input tokens, $10.40 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Windsurf plan is separate.
Do I need an OpenAI account to use GPT-4.1 in Windsurf?
No. Spiral runs GPT-4.1 on its own OpenAI account. You sign in to Spiral once and never paste a key.
How long does it take to set up GPT-4.1 in Windsurf?
About two minutes: add the server, then sign in, then check it.
Does Windsurf wait while GPT-4.1 runs?
GPT-4.1 runs in the background. Windsurf starts the run, checks on it and gives you the text, json when it's ready.
What is GPT-4.1 good at?
GPT-4.1 is useful for code editing, code review, and tasks with detailed response requirements. Its training includes following code diff formats, which makes focused patches a practical use.
How much can GPT-4.1 read and write?
The context window is 1,047,576 tokens, while one response can contain at most 32,768 output tokens. A large input allowance does not mean it can reproduce an equally large document in one answer.
Where does GPT-4.1 struggle with long documents?
Retrieving many separate details from a long context can reduce accuracy. For exhaustive extraction, split the task into smaller sections and check the combined results against the source.
How should I prompt GPT-4.1?
State the task, required format, and exceptions explicitly because GPT-4.1 interprets instructions literally. Specify what it should return when information is missing.
Does valid JSON mean GPT-4.1 extracted the right answer?
No. Structured Outputs can enforce the requested schema, but the values can still be wrong; check extracted facts against the original text.
Start using GPT-4.1 in Windsurf
One Spiral account for every tool your agent needs. No subscription; you pay per run.
Paste this into Windsurf:
set up https://spiral.ai/SKILL.md


