logo
Top up $5, save $2First top-up offer
Live · 2026-11-09

Claude Fable 5.2 API

For complex coding, deep research and agents tackling long, multi-step tasks.

Get API keyDocs

Pay as you go · No Anthropic account needed · Works in Claude Code

Get $5 credit. Pay $3.Save $2 on your first top-up

Built for the whole task

1M context

The standard rate applies across the full context window.

Up to 128K output tokens

Carry a long analysis through to a complete deliverable.

Text + image input

Work with code, documents, screenshots and diagrams.

Adaptive thinking

Use effort to choose how deeply the model reasons.

Work out your API cost

Monthly cost calculator

95%

Estimated monthly costFable 5.2
Anthropic$85.90
BetterToken$58.41

You save $27.49 per month

Get API key

Prices per 1M tokens. Input includes cache hits; cache writes excluded. Sources: BetterToken snapshot and Anthropic pricing, 10 Nov 2026.

Top up once. Get more from every token.

Choose how much to add to your balance. The discount reduces what you pay.

Effective price per 1M tokens after this offer
Added to your balanceYou payYou saveFable 5.2 input / output
$5First top-upYou pay: $3 · You save: $2$3−$2$3.26 / $16.32Top up
$50You pay: $48 · You save: $2$48−$2$5.22 / $26.11Top up
$100You pay: $95 · You save: $5$95−$5$5.17 / $25.84Top up
$200You pay: $185 · You save: $15$185−$15$5.03 / $25.16Top up
$500Lowest tier rateYou pay: $450 · You save: $50$450−$50$4.90 / $24.48Top up

More Claude models. The same API key.

The same balance also works with GPT, Gemini, DeepSeek and more.

per million tokens · Input / Output
Claude modelOfficial input / outputOur input / outputSavings
Fable 5.2$8 / $40$5.44 / $27.20−32%
Fable 5.1$10 / $50$6.80 / $34−32%
Fable 5$10 / $50$6.80 / $34−32%
Opus 5.5$4 / $20$2.72 / $13.60−32%
Opus 5$5 / $25$3.40 / $17−32%
Opus 4.8$5 / $25$3.40 / $17−32%
Opus 4.6$5 / $25$3.40 / $17−32%
Sonnet 5.5$2 / $10$1.36 / $6.80−32%
Sonnet 5$2 / $10$1.36 / $6.80−32%
Sonnet 4.6$3 / $15$2.04 / $10.20−32%
Haiku 5.5from $0.10 / $0.50from $0.07 / $0.34−32%
≤100K prompt: $0.07 / $0.34. >100K: $0.34 / $1.70. Input / output per 1M tokens.
Haiku 4.5$1 / $5$0.68 / $3.40−32%
Explore Claude API →Use Claude Code →

Put Fable 5.2 to work

One base URL and one key for your tools and code.

1. Anthropic Messages

curl https://bettertoken.ai/v1/messages \
  -H "x-api-key: $BETTERTOKEN_API_KEY" \
  -H "anthropic-version: 2023-06-01" \
  -H "content-type: application/json" \
  -d '{
    "model": "claude-fable-5-2",
    "max_tokens": 4096,
    "messages": [{"role": "user", "content": "Hi."}]
  }'

2. Claude Code

Set the endpoint, your BetterToken key and the model, then launch Claude Code.

export ANTHROPIC_BASE_URL="https://bettertoken.ai"
export ANTHROPIC_AUTH_TOKEN="YOUR_BETTERTOKEN_API_KEY"
export ANTHROPIC_MODEL="claude-fable-5-2"
claude
3. OpenAI format
Python
from openai import OpenAI
import os

client = OpenAI(
    api_key=os.environ["BETTERTOKEN_API_KEY"],
    base_url="https://bettertoken.ai/v1",
)
response = client.chat.completions.create(
    model="claude-fable-5-2",
    max_tokens=4096,
    messages=[{"role": "user", "content": "Review this system design."}],
)
print(response.choices[0].message.content)

A few things you might want to know

BetterToken charges $5.44 for input, $27.20 for output and $0.14 for cache reads per million tokens. Anthropic’s standard input / output rates are $8 / $40. Cache writes are billed separately.

Create a BetterToken account or sign in, open Workspace and create an API key. Top up your balance, then use claude-fable-5-2 with the examples above.

Official input / output prices fall from $10 / $50 to $8 / $40 per million tokens, a 20% reduction. On BetterToken they fall from $6.80 / $34 to $5.44 / $27.20. Fable 5.2 is built for demanding reasoning and long-running agent work.

Choose Fable 5.2 for demanding reasoning and long-horizon agent work; choose Opus 5.5 for long-running agentic coding and knowledge work. On BetterToken, Opus 5.5 input / output costs $2.72 / $13.60 per million tokens, versus $5.44 / $27.20 for Fable 5.2. Test with your own workload.

Change the model to claude-fable-5-2. Omit temperature, top_p and top_k. Forced tool_choice values any and tool return 400. Adaptive thinking is always on; use effort to control depth. Anthropic requires 30-day data retention for this model.

You can configure Claude Code with your BetterToken API key and pay for API tokens from your balance. API usage is billed separately from a Claude subscription and remains subject to model and account limits.

The public model catalog lists Anthropic Messages and OpenAI formats. Configure the base URL in the Anthropic Python or Node SDK, or use the curl examples above.

Top up in US dollars with Stripe or USDT on TRC-20. Check current payment options in Workspace. The first top-up gives $5 credit for $3; regular tier offers are listed above.

YOUR NEXT PROJECT STARTS HERE

Give your next big task to Fable 5.2.

Your first $5 of API credit costs $3.

Get API keySee top-up tiers