# Claude Fable 5.2 MCP: setup and a real run with costs

Claude Fable 5.2 MCP setup for Claude Code, claude.ai and the Claude API, plus a real 69-call looot run with the cost of every call.

## Summary

To use Claude Fable 5.2 with MCP, add an MCP server such as looot (https://api.looot.ai/mcp) to Claude Code, claude.ai or the Claude API, then pick the model claude-fable-5-2. In our run on 2026-11-09, Fable 5.2 made 69 looot calls to find 10 leads, and the average cost per call ran from a fraction of a cent for a page of Google results to about two cents for a people search.

Anthropic released Fable 5.2 the same day. It is the model Anthropic points at long, many-step agent work, and a lead hunt with dozens of lookups is that kind of job. Below are the three setups, the receipts from that run, and when Opus 5.5 is the better pick. The full walkthrough, with the exact prompt and every lead, is in the [docs guide](https://docs.looot.ai/guides/claude-fable-5-2-mcp). The model choice gets its own page in [Fable 5.2 vs Opus 5.5 for agent data work](https://docs.looot.ai/guides/claude-fable-5-2-vs-opus-5-5).

## How do I add an MCP server to Claude Code with Fable 5.2?

Add the server once, sign in on looot.ai in the browser, then start Claude Code on Fable 5.2. Three commands. The server is added for every project, so you do it once per machine.

Type /mcp and check that looot shows Connected. Then ask for your balance. That call is free and proves the connection reaches your organization. On a server with no browser, pass an agent token as an Authorization header instead, as the [Claude Code connect page](https://docs.looot.ai/connect/claude-code) shows.

## How do I use Fable 5.2 with MCP in claude.ai?

One custom connector covers claude.ai, Claude Desktop and the Claude mobile apps. You sign in on looot.ai, so there is no token to paste.

1. Open Settings, then Connectors. Click Add, then Add custom connector.
2. Name it looot and set the MCP server URL to https://api.looot.ai/mcp. Leave Advanced settings empty.
3. Click Continue, sign in on looot.ai, pick the organization and click Connect.
4. In a chat, open the + menu, then Connectors, and turn looot on. Pick Claude Fable 5.2 in the model menu.

On Team and Enterprise plans, an Owner adds the connector once for the organization and each member clicks Connect. The [claude.ai connect page](https://docs.looot.ai/connect/claude-ai) has screenshots.

## How do I call claude-fable-5-2 with an MCP server from the Claude API?

Use the MCP connector on the Messages API. It takes two parts: an mcp_servers entry with the URL, and an mcp_toolset tool with the same name, sent with the beta flag mcp-client-2025-11-20. Anthropic's cloud opens the connection, so browser sign-in does not apply. Create an agent token in your looot settings, keep it in an environment variable, and pass it as authorization_token.

Check stop_reason before you read the answer. Fable 5.2 thinks on every request, so a hard task can run for minutes: give the client a long timeout, or stream. The [headless setup page](https://docs.looot.ai/connect/headless) covers tokens for CI and scripts.

## What did a real Claude Fable 5.2 MCP run cost per call?

We gave Claude Code on claude-fable-5-2 one task and no tools except looot: find 10 leads at firms that sell open-source business software, a founder or CEO at each, that person's work email checked, and the firm enriched. It could spend at most one dollar of looot credit. It ran 69 jobs over 74 turns in about 15 minutes and stopped well under the cap. Here is what each job cost.

What each looot job cost in the Claude Fable 5.2 run. Read on 2026-11-09 from [docs.looot.ai/guides/claude-fable-5-2-mcp](https://docs.looot.ai/guides/claude-fable-5-2-mcp).

| Job | Runs | Answered | Charged | Per call |
|---|---|---|---|---|
| google.serp.organic | 20 | 18 (2 provider timeouts, $0) | $0.0073 | $0.000365 on average; $0.0004 or $0.0005 per results page (AnyAPI) |
| people.search | 18 | 17 | $0.360614 | $0.020 on average; AI Ark missed every time and billed $0.016173, then LeadMagic found the person for $0.00198 or $0.00297 |
| people.email.find | 11 | 10 | $0.045 | $0.0041 on average; $0.005 per email found (Kitt AI or Filtrat), $0 on a miss |
| people.email.verify | 10 | 5 valid, 5 catch-all | $0.00125 | $0.000125 on average; $0.00025 per valid answer, $0 for a catch-all answer (Filtrat) |
| company.enrich | 10 | 10 | $0.028 | $0.0028 on average; $0.0019 (The Companies API) or $0.0049 (Hunter) |
| All jobs | 69 | 10 leads | $0.442164 | about $0.044 per lead |

One Claude Code session on claude-fable-5-2, asked for 10 leads with a cap of one dollar of looot credit. Charged is what looot billed for all runs of that job, misses included. Per call is the charged total divided by the runs, then what one answer cost from the provider that gave it.

The people search carried most of the bill. The first provider in line, AI Ark, missed all 18 searches and still billed for each one, which was about two thirds of the data spend. LeadMagic then found the person on 17 of them. The email finder shows the other kind of miss: when Kitt AI found nothing, it charged nothing, and Filtrat answered. Read route.attempts on your own runs to see the same split.

Of the 10 leads, 5 ended with a verified work email and 5 sit on catch-all domains, where the verifier cannot confirm the mailbox. Treat those 5 as unconfirmed. Here is one lead from start to finish, with its run ids:

One lead, run by run: the Zammad CEO. Read on 2026-11-09 from [docs.looot.ai/guides/claude-fable-5-2-mcp](https://docs.looot.ai/guides/claude-fable-5-2-mcp).

| Step | Run id | What happened | Charged |
|---|---|---|---|
| people.search | run_33218e401c5a4036ba1ac3b3b8e5b33c | AI Ark missed and billed $0.016173, then LeadMagic found the CEO for $0.00297 | $0.019143 |
| people.email.find | run_6382068c06d64f0f8526bba4fa209edb | Kitt AI missed at $0, then Filtrat found the email | $0.005 |
| people.email.verify | run_7ed5e5fd11b2462f8753d02353490535 | Filtrat said valid | $0.00025 |
| company.enrich | run_08aa09bd58494725831ff61ffb0587f4 | The Companies API returned the company | $0.0019 |
| One lead, four jobs |  |  | $0.026293 |

Read back after the run with runs_get. Each run id opens the same receipt in your own workspace only if the run is yours; these are ours.

The data side is only part of what the session cost. The model tokens are the other part, and with Fable 5.2 they were the larger one by far.

The whole session: looot data plus Fable 5.2 tokens. Read on 2026-11-09 from [platform.claude.com/docs/en/about-claude/pricing](https://platform.claude.com/docs/en/about-claude/pricing).

| Part | Amount | Cost |
|---|---|---|
| looot data, 69 runs | 10 leads | $0.442164 |
| Fable 5.2 output tokens, $40 per million | 33,175 | $1.33 |
| Fable 5.2 cache reads, $0.20 per million | 1,653,760 | $0.33 |
| Fable 5.2 cache writes (1 hour), $16 per million | 283,977 | $4.54 |
| Fable 5.2 uncached input, $8 per million | 28 | $0.0002 |
| Total |  | $6.64 |

Token counts are what Claude Code reported for the session. The token rows are our arithmetic at Anthropic's list prices for claude-fable-5-2, read 2026-11-09: $8 per million input tokens, $40 per million output tokens, cache reads at 2.5% of input, one-hour cache writes at twice input. The session ran on a Claude subscription, so it counted toward the plan's usage, not an API bill.

Most of the token bill is Claude Code writing the growing conversation to the prompt cache. The looot side would be the same with any other model, since a lookup costs the same whichever model asks for it.

## Should I run my MCP agent on Fable 5.2 or Opus 5.5?

Start with Opus 5.5 unless you have a reason not to. That is Anthropic's own advice: Opus 5.5 for most workloads, Fable 5.2 for demanding reasoning and long-horizon agent work, or when Opus 5.5 at higher effort still falls short ([models overview](https://platform.claude.com/docs/en/about-claude/models/overview), read 2026-11-09).

Fable 5.2 lists at $8 per million input tokens and $40 per million output tokens, twice the $4 and $20 of Opus 5.5 ([Anthropic pricing](https://platform.claude.com/docs/en/about-claude/pricing), read 2026-11-09). Both have a 1M token context window and 128K max output. Fable 5.2 defaults to high effort, Opus 5.5 to medium.

We did not run the same task on Opus 5.5, so the run above shows what one Fable 5.2 session cost, not which model is better at it. The [comparison guide](https://docs.looot.ai/guides/claude-fable-5-2-vs-opus-5-5) works through when the extra token cost pays off. For model prices side by side, see [AI model prices](https://looot.ai/models).

## What looot does and does not do

- **Prepaid, per call.** A workspace starts at zero and spends credit you top up, one call at a time. The [pricing page](https://looot.ai/pricing) explains how a charge is worked out.
- **A price before you spend.** search and inspect are free and show each endpoint's price and the most a run can cost, so the agent can check before it runs.
- **Fallback on a miss.** With fallback, a run that misses tries the next provider for the same job, up to the maxCostUsd you set. A provider that ran and found nothing can still bill, as AI Ark did above.
- **Reads data, does not act as you.** looot does not post on social media, log in to sites as you, buy things or read private accounts.

If you want the same server in other clients, [Claude MCP servers for data work](https://looot.ai/blog/claude-mcp-servers-for-data-work) lists what else Claude can reach, and [one prepaid balance for an MCP server](https://looot.ai/blog/mcp-server-with-one-prepaid-balance) explains the billing model.

## Questions

### Does Claude Fable 5.2 support MCP servers?

Yes. In Claude Code and claude.ai it uses the MCP servers you connect, the same way earlier Claude models do. On the Claude API it reaches remote MCP servers through the MCP connector, with an mcp_servers entry and an mcp_toolset tool.

### What is the model id for Claude Fable 5.2 with MCP in Claude Code?

The model id is claude-fable-5-2. Start Claude Code with claude --model claude-fable-5-2, or pass the same flag with -p for a one-shot run, as our demo did. The MCP server setup does not change with the model.

### How much does a Claude Fable 5.2 MCP run cost?

You pay two bills: model tokens to Anthropic and data calls to the MCP server. In our 10-lead run the tokens were most of the cost, and the looot calls were the smaller part. The receipts tables above give every number, read 2026-11-09.

### Do I need a subscription to use looot with Fable 5.2?

No. looot is prepaid credit spent per call, at the price search shows before the run. When the balance is too low, the run is blocked and nothing is charged.

### Does the looot MCP server work with Opus 5.5 and Sonnet 5.5 too?

Yes. looot is a standard MCP server, so any Claude model that can use MCP tools calls the same tools at the same prices. In Claude Code, pass --model claude-opus-5-5 or --model claude-sonnet-5-5 instead.

### Why did some lookups cost money and find nothing?

Some providers bill for a search that runs and finds nothing, and others do not. In our run AI Ark billed for every miss and Kitt AI did not. Each run's route.attempts lists which provider ran and what it charged, so you can see the split on your own runs.

### Where is the full Claude Fable 5.2 MCP walkthrough?

The [docs guide](https://docs.looot.ai/guides/claude-fable-5-2-mcp) has the exact prompt, the command we ran, all 10 leads and the per-job table. The [comparison guide](https://docs.looot.ai/guides/claude-fable-5-2-vs-opus-5-5) covers when to use Fable 5.2 and when Opus 5.5 is enough.

## For agents

### Code (bash)

```bash
# 1. Add looot for every project
claude mcp add --scope user --transport http looot https://api.looot.ai/mcp

# 2. Sign in: pick the organization to bill and click Connect
claude mcp login looot

# 3. Start a session on Fable 5.2
claude --model claude-fable-5-2
```

### Code (python)

```python
import os
import anthropic

client = anthropic.Anthropic()

response = client.beta.messages.create(
    model="claude-fable-5-2",
    max_tokens=16000,
    betas=["mcp-client-2025-11-20"],
    mcp_servers=[
        {
            "type": "url",
            "url": "https://api.looot.ai/mcp",
            "name": "looot",
            "authorization_token": os.environ["LOOOT_TOKEN"],
        }
    ],
    tools=[{"type": "mcp_toolset", "mcp_server_name": "looot"}],
    messages=[{"role": "user", "content": "Use looot: check my balance, then find and verify the work email of the CEO of example.com."}],
)

for block in response.content:
    if block.type == "text":
        print(block.text)
```

### Try it in Claude Code on Fable 5.2

```text
Use looot to find one lead for me. First set up https://looot.ai/skill.md if looot is not connected yet, and check my balance. Then find the CEO of a company I name, find their work email, verify it, and enrich the company. Tell me the price before each paid call, run each job with fallback, and give me a table of the run ids and what each one charged.
```