How to Use Inkling 2 Mini in Codex
Inkling 2 Mini is a compact reasoning model for coding, research with tools, and document questions. Its strengths are multistep agent tasks; long conversations and instructions with many constraints need closer review.
Prompts to try with Inkling 2 Mini in Codex
Copy an example into Codex, or ask in your own words.
Use Inkling 2 Mini to investigate the failing tests in this project. Reproduce one failure, trace its cause, make a focused fix, and run the relevant tests. Report the cause and the commands that passed.
Use Inkling 2 Mini to inspect the pagination code and its tests. Find one uncovered boundary case, add a test with an explicit expected result, and fix the implementation if that test exposes a bug.
Use Inkling 2 Mini to compare this project's API client with the migration guide in the repository. List affected call sites, update one component, and run its tests before continuing.
Use Inkling 2 Mini to trace how this project handles an expired session. Follow the request from the UI through the server, identify inconsistent behavior, and propose a patch with a regression test.
Use Inkling 2 Mini with medium reasoning effort to review the current diff for behavior changes. Check each finding against the surrounding code and tests, and report only issues with a reproducible example.
How to set up Inkling 2 Mini in Codex
Pick the way that suits you. Each takes about two minutes.
Use Spiral's MCP server
- Add the hosted server
codex mcp add spiral --url https://api.spiral.ai/mcp
- Sign in
Codex opens the browser: sign in to Spiral and click Authorize. No API key. codex mcp list then shows Spiral.
codex mcp login spiral
Use the Spiral skill
- Send Codex 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 Inkling 2 Mini
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": "thinkingmachines/inkling-2-mini", "input": { "messages": [ { "role": "user", "content": "Hello!" } ] } }'
Then paste one of the prompts above: Codex finds Inkling 2 Mini on Spiral, shows the price and runs it.
Inkling 2 Mini providers and prices
Spiral runs Inkling 2 Mini through OpenRouter. Open a card for today's price, with every fee included.
Up to $0.034084 per run at the default inputs. $0.13 per 1M input tokens, $0.52 per 1M output tokens.
Settings you can ask for
Ask Codex for any of these in your prompt.
| Setting | Allowed values | Default | Description |
|---|---|---|---|
| Seed | a whole number | ||
| Min p | 0 to 1 | ||
| Top k | a whole number from 0 | ||
| Top p | 0 to 1 | ||
| Logprobs | on or off | ||
| Max tokens | 1 to 65536 | The most output tokens to generate. On a reasoning model the reasoning tokens count toward it: a cap too low for the effort can end with no answer (finish_reason length). | |
| Temperature | 0 to 2 | ||
| Top logprobs | 0 to 20 | ||
| Presence penalty | -2 to 2 | ||
| Reasoning effort | none, minimal, low, medium, high, or max | high | How much the model reasons. |
| Frequency penalty | -2 to 2 | ||
| Repetition penalty | 0 to 2 |
Every API field of Thinking Machines: Inkling 2 Mini is on its tool page: open its card above.
What can you do with Inkling 2 Mini in Codex?
Inkling 2 Mini can help carry a bug fix or migration through investigation, edits, and checks. Give it the relevant project files and explicit acceptance criteria, then review the patch and test results.
Fix a reproducible bug
Give Inkling 2 Mini the failing behavior, relevant code, and expected result. Ask for a focused patch and a regression test that demonstrates the fix.
Plan and carry out a code migration
Ask Inkling 2 Mini to identify affected call sites, update one component at a time, and check behavior against the migration requirements.
Find missing test coverage
Have Inkling 2 Mini compare an implementation with its existing tests, identify uncovered behavior, and propose cases with explicit expected results.
Research a narrow question
Use Inkling 2 Mini with search tools to investigate a specific claim. Request primary sources, supporting passages, and a clear account of unresolved evidence.
Answer questions across long documents
Provide labeled source text and ask Inkling 2 Mini to locate relevant passages, reconcile conflicting statements, and cite the evidence behind its answer.
Prepare a recurring change brief
Have Inkling 2 Mini compare newly retrieved announcements with the previous run's summary. Request only substantive changes, source links, and open questions.
Turn support reports into investigation steps
Give Inkling 2 Mini a set of reports and ask it to separate observed symptoms from suspected causes, group related issues, and propose reproducible checks.
More on Inkling 2 Mini and Codex
Overview
Inkling 2 Mini in other agents
Questions about Inkling 2 Mini in Codex
Can Codex use Inkling 2 Mini?
Yes. Add Spiral to Codex as an MCP server, and Codex can find and run Inkling 2 Mini like any built-in tool.
How much does Inkling 2 Mini cost in Codex?
Up to $0.034084 per run at the default inputs. $0.13 per 1M input tokens, $0.52 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Codex plan is separate.
Do I need an OpenRouter account to use Inkling 2 Mini in Codex?
No. Spiral runs Inkling 2 Mini on its own OpenRouter account. You sign in to Spiral once and never paste a key.
How long does it take to set up Inkling 2 Mini in Codex?
About two minutes: add the hosted server, then sign in.
Does Codex wait while Inkling 2 Mini runs?
Inkling 2 Mini runs in the background. Codex starts the run, checks on it and gives you the text, json when it's ready.
What is Inkling 2 Mini best suited for?
Coding tasks that require several steps and research that uses search or other tools are good starting points. Give it a concrete goal and a way to check the result, such as a failing test or a source citation.
Where does Inkling 2 Mini need closer review?
Thinking Machines flags long conversations and instructions with many constraints as weaker areas. Keep the current task self-contained, list the essential requirements, and check the result against them.
How much context can Inkling 2 Mini handle?
The published model supports up to one million tokens of context. For document questions, label the sources and ask for supporting passages so you can check each conclusion.
Can Inkling 2 Mini understand or generate images?
It accepts images alongside text and returns text. For optimal image performance, Thinking Machines recommends keeping each image dimension between 40 and 4,096 pixels.
Does Inkling 2 Mini support languages besides English?
Yes. Its model card lists English and general multilingual support; specify the output language and review terminology for your audience.
Start using Inkling 2 Mini in Codex
One Spiral account for every tool your agent needs. No subscription; you pay per run.
Paste this into Codex:
set up https://spiral.ai/SKILL.md


