[Integrations](https://spiral.ai/integrations) › [Codex](https://spiral.ai/integrations/codex) › [Claude Fable 5.2](https://spiral.ai/integrations/codex/claude-fable-5-2)

# How to Use Claude Fable 5.2 in Codex

Claude Fable 5.2 is built for difficult coding, deep research, and extended technical tasks. Compared with Claude Fable 5.1, Anthropic reports improvements in coding, reasoning, and vision.

|  |  |
| --- | --- |
| Model | Claude Fable 5.2 |
| Available through | Anthropic |
| Category | LLM |
| Price | Up to $0.21372 per run at the default inputs. $10.40 per 1M input tokens, $0.26 per 1M cached input tokens, $13.00 per 1M cache write tokens, $20.80 per 1M 1-hour cache write tokens, $52.00 per 1M output tokens. |
| Returns | text, json |
| Runs | In the background |
| Setup | About 2 minutes, no API key |

## Prompts to try with Claude Fable 5.2 in Codex

- Use Claude Fable 5.2 to investigate the failing tests in this repository at high reasoning effort. Reproduce the failure, trace its cause, make a focused fix, and run the relevant checks. Finish with the cause, changes, results, and any remaining blocker.
- Use Claude Fable 5.2 to review this project's background job retry logic. Identify paths that could repeat a completed operation, explain the evidence, and propose the smallest correction. Keep the final report under 600 words.
- Use Claude Fable 5.2 to plan the database migration described in this project's issue notes. Inspect the schema and callers, identify compatibility risks, and produce an ordered implementation plan with rollback criteria and acceptance checks.
- Use Claude Fable 5.2 to compare the two caching approaches documented in this repository. Evaluate invalidation behavior, concurrency, and failure recovery against the project's requirements. Recommend one approach and identify assumptions that need testing.
- Use Claude Fable 5.2 to complete the refactor described in the project task. Preserve documented behavior, verify the affected paths, and continue through recoverable errors until the acceptance checks pass or a concrete blocker remains. Report the actual diff and check results.

## How to set up Claude Fable 5.2 in Codex

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

### Use Spiral's MCP server

1. **Add the hosted server**

```bash
codex mcp add spiral --url https://api.spiral.ai/mcp
```

2. **Sign in**: Codex opens the browser: sign in to Spiral and click Authorize. No API key. `codex mcp list` then shows Spiral.

```bash
codex mcp login spiral
```

### Use the Spiral skill

1. **Send Codex 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 Claude Fable 5.2**: 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": "anthropic",
    "model": "anthropic/claude-fable-5.2",
    "input": {
      "messages": [
        {
          "role": "system"
        }
      ]
    }
  }'
```

Then paste one of the prompts above: Codex finds Claude Fable 5.2 on Spiral, shows the price and runs it.

Setup guide in the docs: https://spiral.ai/docs/quickstarts/codex

## Claude Fable 5.2 tools and prices

- [Claude Fable 5.2](https://spiral.ai/tools/anthropic/anthropic/claude-fable-5.2) (Anthropic): Up to $0.21372 per run at the default inputs. $10.40 per 1M input tokens, $0.26 per 1M cached input tokens, $13.00 per 1M cache write tokens, $20.80 per 1M 1-hour cache write tokens, $52.00 per 1M output tokens.

## Settings you can ask for

| Setting | Allowed values | Default | Description |
| --- | --- | --- | --- |
| stop |  | — |  |
| tools |  | — |  |
| reasoning |  | — | Reasoning effort; reasoning tokens count toward max_tokens. |
| max_tokens | 1 to 128000 | — | 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). |
| tool_choice |  | — |  |
| response_format |  | — |  |

## What can you do with Claude Fable 5.2 in Codex?

Turn difficult code investigations into reviewed changes with Claude Fable 5.2. Give the agent a defined scope, acceptance checks, and a final report that distinguishes completed work from unresolved issues.

### Diagnose a difficult software bug

Ask Claude Fable 5.2 to trace a failure through the relevant code, identify a reproducible cause, and propose a focused fix with verification.

### Plan a complex code change

Have Claude Fable 5.2 map dependencies, identify behavior that must stay consistent, and develop an implementation plan with acceptance checks.

### Compare technical approaches

Give Claude Fable 5.2 competing designs and explicit requirements. Request a recommendation that identifies assumptions, tradeoffs, and unanswered questions.

### Synthesize technical research

Provide source excerpts and ask Claude Fable 5.2 to reconcile competing findings, distinguish evidence from inference, and identify the next questions to investigate.

### Turn analysis into a technical report

Ask Claude Fable 5.2 to develop supplied findings into a finished report with recommendations, supporting evidence, and clearly stated gaps.

### Review a multistep technical investigation

Have Claude Fable 5.2 assess completed experiments, identify unresolved hypotheses, and produce a prioritized plan for the remaining investigation.

## Overview

- [Anthropic Claude](https://spiral.ai/providers/anthropic)
- [All Codex guides](https://spiral.ai/integrations/codex)

## Claude Fable 5.2 in other agents

- [Amp](https://spiral.ai/integrations/amp/claude-fable-5-2)
- [ChatGPT](https://spiral.ai/integrations/chatgpt/claude-fable-5-2)
- [Claude](https://spiral.ai/integrations/claude-desktop/claude-fable-5-2)
- [Claude Code](https://spiral.ai/integrations/claude-code/claude-fable-5-2)
- [Cline](https://spiral.ai/integrations/cline/claude-fable-5-2)
- [Cursor](https://spiral.ai/integrations/cursor/claude-fable-5-2)
- [Devin](https://spiral.ai/integrations/devin/claude-fable-5-2)
- [dots](https://spiral.ai/integrations/dots/claude-fable-5-2)
- [Factory Droid](https://spiral.ai/integrations/factory-droid/claude-fable-5-2)
- [Gemini](https://spiral.ai/integrations/gemini/claude-fable-5-2)
- [Gemini CLI](https://spiral.ai/integrations/gemini-cli/claude-fable-5-2)
- [Google Antigravity](https://spiral.ai/integrations/antigravity/claude-fable-5-2)
- [goose](https://spiral.ai/integrations/goose/claude-fable-5-2)
- [Grok](https://spiral.ai/integrations/grok/claude-fable-5-2)
- [Grok Build](https://spiral.ai/integrations/grok-build/claude-fable-5-2)
- [Hermes](https://spiral.ai/integrations/hermes/claude-fable-5-2)
- [Instinct](https://spiral.ai/integrations/instinct/claude-fable-5-2)
- [Manus](https://spiral.ai/integrations/manus/claude-fable-5-2)
- [Muse](https://spiral.ai/integrations/muse/claude-fable-5-2)
- [Muse Code](https://spiral.ai/integrations/muse-code/claude-fable-5-2)
- [n8n](https://spiral.ai/integrations/n8n/claude-fable-5-2)
- [Notion](https://spiral.ai/integrations/notion/claude-fable-5-2)
- [OpenClaw](https://spiral.ai/integrations/openclaw/claude-fable-5-2)
- [opencode](https://spiral.ai/integrations/opencode/claude-fable-5-2)
- [Perplexity](https://spiral.ai/integrations/perplexity/claude-fable-5-2)
- [Raycast](https://spiral.ai/integrations/raycast/claude-fable-5-2)
- [VS Code](https://spiral.ai/integrations/vscode/claude-fable-5-2)
- [Windsurf](https://spiral.ai/integrations/windsurf/claude-fable-5-2)
- [Zapier](https://spiral.ai/integrations/zapier/claude-fable-5-2)

## More Codex guides

- [Claude Fable 5.1](https://spiral.ai/integrations/codex/claude-fable-5-1)
- [Claude Fable Latest](https://spiral.ai/integrations/codex/claude-fable-latest)
- [Claude Fable 5](https://spiral.ai/integrations/codex/claude-fable-5)
- [Claude Haiku 4.5](https://spiral.ai/integrations/codex/claude-haiku-4-5)
- [Beam 1.0 501B A23B](https://spiral.ai/integrations/codex/beam-1-0-501b-a23b)
- [Claude Haiku 5.5](https://spiral.ai/integrations/codex/claude-haiku-5-5)

## Questions about Claude Fable 5.2 in Codex

**Can Codex use Claude Fable 5.2?**

Yes. Add Spiral to Codex as an MCP server, and Codex can find and run Claude Fable 5.2 like any built-in tool.

**How much does Claude Fable 5.2 cost in Codex?**

Up to $0.21372 per run at the default inputs. $10.40 per 1M input tokens, $0.26 per 1M cached input tokens, $13.00 per 1M cache write tokens, $20.80 per 1M 1-hour cache write tokens, $52.00 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 Anthropic account to use Claude Fable 5.2 in Codex?**

No. Spiral runs Claude Fable 5.2 on its own Anthropic account. You sign in to Spiral once and never paste a key.

**How long does it take to set up Claude Fable 5.2 in Codex?**

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

**Does Codex wait while Claude Fable 5.2 runs?**

Claude Fable 5.2 runs in the background. Codex starts the run, checks on it and gives you the text, json when it's ready.

**What is Claude Fable 5.2 best suited for?**

Use it for difficult coding, multistep research, and technical analysis that needs a finished deliverable. Give it a concrete result to produce and clear criteria for completion.

**How much context and output can Claude Fable 5.2 handle?**

It supports a one-million-token context window and up to 128,000 output tokens. These are token limits, so avoid treating them as fixed page or word counts.

**Which reasoning effort should I use?**

Start with high effort and compare results on representative tasks. Increase effort for difficult work only when testing shows a benefit; lower it when the results still meet your requirements.

**Can I turn Claude Fable 5.2 thinking off?**

Adaptive thinking is always enabled. Adjust effort to control thinking depth, and request a specific answer length separately.

**What should I do if Claude Fable 5.2 stops before finishing?**

State the remaining deliverable and tell it to continue through ordinary, recoverable obstacles. Early stopping is more common at lower effort levels, so test high effort for extended tasks.

