How to Use Devstral 2 in ChatGPT
Devstral 2 is built for bug fixes and coordinated changes across multiple files. Its 256K context window suits coding tasks that need substantial repository context.
Prompts to try with Devstral 2 in ChatGPT
Copy an example into ChatGPT, or ask in your own words.
Use Devstral 2 to review the code and stack trace pasted in this conversation. Explain the most likely failure path, propose a minimal correction, and give a test that would distinguish your diagnosis from alternatives.
Use Devstral 2 to explain the pasted request handler and database code. Trace one successful request and one rejected request, then identify any mismatch between validation and persistence.
Use Devstral 2 to refactor the pasted functions that duplicate date parsing. Preserve their documented behavior and show tests for invalid dates, empty input, and time zone boundaries.
Use Devstral 2 to compare the two implementations pasted above. Focus on behavior, error handling, and compatibility; support each finding with a specific code excerpt.
Use Devstral 2 with randomness set to 0.2 to turn the pasted bug report into a debugging plan. List what the evidence establishes, what remains uncertain, and the next experiment for each uncertainty.
How to set up Devstral 2 in ChatGPT
Pick the way that suits you. Each takes about two minutes.
Use the ChatGPT connector
- Add a custom MCP server
In ChatGPT open Plugins, then Add › Add custom MCP server. Name it Spiral, paste this URL and keep Authentication on OAuth. Tick I understand and want to continue, then click Create.
https://api.spiral.ai/mcp
- Click Connect
ChatGPT asks to connect Spiral. Connect takes you to Spiral in your browser.
- Sign in to Spiral and approve
Sign in, pick the organization ChatGPT should use and click Authorize. Spiral gives the connection its own API key; ChatGPT never sees it, and you can remove the connection under Settings › Connected apps.
Use the Spiral skill
- Send ChatGPT 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 Devstral 2
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": "openrouter", "model": "mistralai/devstral-2512", "input": { "messages": [ { "role": "system" } ] } }'
Then paste one of the prompts above: ChatGPT finds Devstral 2 on Spiral, shows the price and runs it.
Devstral 2 providers and prices
Spiral runs Devstral 2 through OpenRouter. Open a card for today's price, with every fee included.
Up to $0.010668 per run at the default inputs. $0.52 per 1M input tokens, $2.60 per 1M output tokens.
Settings you can ask for
Ask ChatGPT for any of these in your prompt.
| Setting | Allowed values | Default | Description |
|---|---|---|---|
| Seed | a whole number | ||
| Top p | 0 to 1 | ||
| Max tokens | 1 to 209715 | The most output tokens to generate. | |
| Temperature | 0 to 2 | ||
| Presence penalty | -2 to 2 | ||
| Frequency penalty | -2 to 2 |
Every API field of Mistral: Devstral 2 2512 is on its tool page: open its card above.
What can you do with Devstral 2 in ChatGPT?
Devstral 2 can turn pasted code and error output into a focused debugging discussion. Ask for explanations and proposed changes tied to the evidence you provide.
Fix bugs across application layers
Use Devstral 2 to trace a reported failure through validation, business logic, and persistence, then propose a focused fix with a regression test.
Refactor a shared interface
Ask Devstral 2 to update an interface and its callers together, preserving existing behavior and identifying tests affected by the change.
Plan a legacy code migration
Give Devstral 2 the current implementation and target requirements. Request incremental changes, compatibility checks, and a rollback plan.
Explain an unfamiliar codebase
Use Devstral 2 to trace a request through supplied code and explain the responsibilities, dependencies, and failure paths along the way.
Investigate failing tests
Provide Devstral 2 with failure output and relevant code. Ask for the likely cause, the evidence supporting it, and a testable correction.
Review a coordinated code change
Ask Devstral 2 to review related changes together, checking whether callers, validation rules, and tests still agree on the expected behavior.
More on Devstral 2 and ChatGPT
Overview
Devstral 2 in other agents
Questions about Devstral 2 in ChatGPT
Can ChatGPT use Devstral 2?
Yes. Add Spiral to ChatGPT as an MCP server, and ChatGPT can find and run Devstral 2 like any built-in tool.
How much does Devstral 2 cost in ChatGPT?
Up to $0.010668 per run at the default inputs. $0.52 per 1M input tokens, $2.60 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your ChatGPT plan is separate.
Do I need an OpenRouter account to use Devstral 2 in ChatGPT?
No. Spiral runs Devstral 2 on its own OpenRouter account. You sign in to Spiral once and never paste a key.
How long does it take to set up Devstral 2 in ChatGPT?
About two minutes: add a custom MCP server, then click Connect, then sign in to Spiral and approve.
Does ChatGPT wait while Devstral 2 runs?
Devstral 2 runs in the background. ChatGPT starts the run, checks on it and gives you the text, json when it's ready.
What is Devstral 2 best suited for?
Devstral 2 targets software engineering: exploring codebases, fixing bugs, and coordinating edits across multiple files. Use it for tasks that require understanding how several parts of an application fit together.
How much context can Devstral 2 handle?
Mistral specifies a 256K context window. Treat that as a context limit, not a promise that one response can contain 256K tokens of generated code.
What temperature should I use for coding?
Mistral recommends 0.2. Use that as a starting point when evaluating bug fixes and code changes against your own tests.
How should I prompt Devstral 2?
Give it a concrete task, relevant code, expected behavior, and an example of the desired result. For a long change, ask it to divide the work into steps and report how each step was checked.
Where does Devstral 2 fall short?
Mistral reports that Claude Sonnet 4.5 remained significantly preferred in its human evaluation through Cline. Review proposed changes and test them against the behavior you need.
Start using Devstral 2 in ChatGPT
One Spiral account for every tool your agent needs. No subscription; you pay per run.
Paste this into ChatGPT:
set up https://spiral.ai/SKILL.md


