[Integrations](https://spiral.ai/integrations) › [Claude Code](https://spiral.ai/integrations/claude-code) › [Pareto 26.9](https://spiral.ai/integrations/claude-code/pareto-26-9)

# How to Use Pareto 26.9 in Claude Code

Use Pareto 26.9 for code fixes, document questions, and extracting records from text. Its 262,144-token context supports substantial source material, but JSON output still needs validation against your required structure.

|  |  |
| --- | --- |
| Model | Pareto 26.9 |
| Available through | OpenRouter |
| Category | LLM |
| Price | Up to $0.040051 per run at the default inputs. $3.25 per 1M input tokens, $9.75 per 1M output tokens. |
| Returns | text, json |
| Runs | In the background |
| Setup | About 2 minutes, no API key |

## Prompts to try with Pareto 26.9 in Claude Code

- Use Pareto 26.9 to read the failing test output and relevant project files, identify the cause, and propose the smallest code fix. Keep the existing assertions and rerun the affected tests.
- Use Pareto 26.9 to review the current diff for correctness bugs. For each finding, give the file, line, triggering input, and expected versus actual behavior.
- Use Pareto 26.9 to inspect our support-ticket extraction code and sample messages. Propose explicit category rules and represent missing order numbers as null.
- Use Pareto 26.9 to answer how this project's retry behavior works from its code and documentation. Cite file locations and identify any disagreement between the implementation and the docs.
- Use Pareto 26.9 to review our JSON record validator against the extraction requirements. Check required fields, types, allowed categories, and missing-value handling, then report gaps with examples.

## How to set up Pareto 26.9 in Claude Code

Pick the way that suits you. Each takes about two minutes.

### Use Spiral's MCP server

1. **Add the hosted server**: This adds Spiral to the current project. Add `--scope user` to use it in every project.

```bash
claude mcp add --transport http spiral https://api.spiral.ai/mcp
```

2. **Sign in**: Start `claude`, type `/mcp`, choose spiral and Authenticate. Sign in to Spiral in the browser and click Authorize. No API key.

### Use the Spiral skill

1. **Send Claude Code this message**: Your agent reads Spiral's skill and sets itself up.

```
set up https://spiral.ai/SKILL.md
```

2. **Sign in or paste a key**: Approve Spiral in the browser, or give the agent a key from the console.

### Use the Spiral CLI

1. **Install and sign in**: Installs the Spiral CLI, signs you in and adds its skill to your coding agents.

```bash
npx -y spiral-cli@latest init
```

### Call the Spiral API

1. **Create an API key**: In the console, under API keys.

2. **Run Pareto 26.9**: One request; the result comes back by polling or a webhook.

```bash
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": "unbiased/pareto-26.9",
    "input": {
      "messages": [
        {
          "role": "system"
        }
      ]
    }
  }'
```

Then paste one of the prompts above: Claude Code finds Pareto 26.9 on Spiral, shows the price and runs it.

Setup guide in the docs: https://spiral.ai/docs/quickstarts/claude-code

## Pareto 26.9 tools and prices

- [Pareto 26.9](https://spiral.ai/tools/openrouter/unbiased/pareto-26.9) (OpenRouter): Up to $0.040051 per run at the default inputs. $3.25 per 1M input tokens, $9.75 per 1M output tokens.

## Settings you can ask for

| Setting | Allowed values | Default | Description |
| --- | --- | --- | --- |
| tools |  | — |  |
| top_p | 0 to 1 | — |  |
| max_tokens | 1 to 131072 | — | The most output tokens to generate. |
| temperature | 0 to 2 | — |  |
| tool_choice |  | — |  |
| response_format |  | — |  |

## What can you do with Pareto 26.9 in Claude Code?

Turn failing tests and pull request diffs into proposed fixes and review findings with Pareto 26.9. Give the agent relevant project files and a concrete check for each requested change.

### Propose a regression fix

Give Pareto 26.9 the failing test output and relevant code. Ask for a focused patch, then rerun the tests to check the result.

### Review a pull request

Ask Pareto 26.9 to examine a diff for correctness problems. Require a location, a failure scenario, and supporting code for each finding.

### Extract support records

Turn support messages into draft records containing customer names, order numbers, categories, and amounts. Ask Pareto 26.9 to leave unstated values null, then validate the records.

### Classify support requests

Give Pareto 26.9 explicit category definitions and urgency rules. Request one label and a supporting passage for each message.

### Answer questions from documentation

Supply relevant documentation and ask Pareto 26.9 for an answer tied to specific passages. Require it to identify questions the supplied text does not answer.

### Extract records from operational notes

Ask Pareto 26.9 to turn freeform notes into records with explicitly defined fields. Preserve source excerpts so a reviewer can check each extracted value.

## Overview

- [OpenRouter](https://spiral.ai/providers/openrouter)
- [All Claude Code guides](https://spiral.ai/integrations/claude-code)

## Pareto 26.9 in other agents

- [Amp](https://spiral.ai/integrations/amp/pareto-26-9)
- [ChatGPT](https://spiral.ai/integrations/chatgpt/pareto-26-9)
- [Claude](https://spiral.ai/integrations/claude-desktop/pareto-26-9)
- [Cline](https://spiral.ai/integrations/cline/pareto-26-9)
- [Codex](https://spiral.ai/integrations/codex/pareto-26-9)
- [Cursor](https://spiral.ai/integrations/cursor/pareto-26-9)
- [Devin](https://spiral.ai/integrations/devin/pareto-26-9)
- [dots](https://spiral.ai/integrations/dots/pareto-26-9)
- [Factory Droid](https://spiral.ai/integrations/factory-droid/pareto-26-9)
- [Gemini](https://spiral.ai/integrations/gemini/pareto-26-9)
- [Gemini CLI](https://spiral.ai/integrations/gemini-cli/pareto-26-9)
- [Google Antigravity](https://spiral.ai/integrations/antigravity/pareto-26-9)
- [goose](https://spiral.ai/integrations/goose/pareto-26-9)
- [Grok](https://spiral.ai/integrations/grok/pareto-26-9)
- [Grok Build](https://spiral.ai/integrations/grok-build/pareto-26-9)
- [Hermes](https://spiral.ai/integrations/hermes/pareto-26-9)
- [Instinct](https://spiral.ai/integrations/instinct/pareto-26-9)
- [Manus](https://spiral.ai/integrations/manus/pareto-26-9)
- [Muse](https://spiral.ai/integrations/muse/pareto-26-9)
- [Muse Code](https://spiral.ai/integrations/muse-code/pareto-26-9)
- [n8n](https://spiral.ai/integrations/n8n/pareto-26-9)
- [Notion](https://spiral.ai/integrations/notion/pareto-26-9)
- [OpenClaw](https://spiral.ai/integrations/openclaw/pareto-26-9)
- [opencode](https://spiral.ai/integrations/opencode/pareto-26-9)
- [Perplexity](https://spiral.ai/integrations/perplexity/pareto-26-9)
- [Raycast](https://spiral.ai/integrations/raycast/pareto-26-9)
- [VS Code](https://spiral.ai/integrations/vscode/pareto-26-9)
- [Windsurf](https://spiral.ai/integrations/windsurf/pareto-26-9)
- [Zapier](https://spiral.ai/integrations/zapier/pareto-26-9)

## More Claude Code guides

- [Pareto 26.10 Preview](https://spiral.ai/integrations/claude-code/pareto-26-10-preview)
- [Pareto Latest](https://spiral.ai/integrations/claude-code/pareto-latest)
- [Pareto 26.10](https://spiral.ai/integrations/claude-code/pareto-26-10)
- [Particle-1](https://spiral.ai/integrations/claude-code/particle-1)
- [Palmyra X5](https://spiral.ai/integrations/claude-code/palmyra-x5)
- [Perceptron Mk1](https://spiral.ai/integrations/claude-code/perceptron-mk1)

## Questions about Pareto 26.9 in Claude Code

**Can Claude Code use Pareto 26.9?**

Yes. Add Spiral to Claude Code as an MCP server, and Claude Code can find and run Pareto 26.9 like any built-in tool.

**How much does Pareto 26.9 cost in Claude Code?**

Up to $0.040051 per run at the default inputs. $3.25 per 1M input tokens, $9.75 per 1M output tokens. You pay from your Spiral balance as you go, with no subscription, and a failed run is not charged. Your Claude Code plan is separate.

**Do I need an OpenRouter account to use Pareto 26.9 in Claude Code?**

No. Spiral runs Pareto 26.9 on its own OpenRouter account. You sign in to Spiral once and never paste a key.

**How long does it take to set up Pareto 26.9 in Claude Code?**

About two minutes: add the hosted server, then sign in.

**Does Claude Code wait while Pareto 26.9 runs?**

Pareto 26.9 runs in the background. Claude Code starts the run, checks on it and gives you the text, json when it's ready.

**What should I try Pareto 26.9 on first?**

Start with a bounded task you can check: a failing test, a pull request diff, or records extracted from text. Define what a correct result looks like before comparing answers.

**How much text can Pareto 26.9 handle?**

Its context window is 262,144 tokens, with up to 131,072 output tokens. Ask for the length your task needs instead of treating the output ceiling as a target.

**Does Pareto 26.9 guarantee that JSON matches my required structure?**

No. JSON output does not enforce a JSON schema, so validate required fields, types, and allowed labels before using the result.

**How should I prompt Pareto 26.9 to extract records?**

Name the information you need, define classification rules, and tell it to return null when the source omits a value. For urgency, give a concrete rule such as an explicitly stated deadline.

**How strong is the evidence for Pareto 26.9's coding performance?**

The maker reports a DeepSWE evaluation, but it used a 30-task slice rather than the full benchmark. Test realistic changes in your own project before relying on a benchmark comparison.

