Breaking

Claude Developers Publishes Guide on Choosing Among Fable 5.2, Opus 5.5, Sonnet and Haiku

November 9, 2026 at 16:13 EST

Foundation Models
AI Agents

The Claude Developers account has published a guide to help builders decide which Claude model to use, covering Fable 5.2, Opus 5.5, Sonnet and Haiku, along with when to adjust effort instead of switching models and how to cut costs by running subagents on smaller models.

Developer guide · Claude model selection

Which Claude model fits your build? A new guide tackles the operational decisions

The guide covers when to use Fable 5.2, Opus 5.5, Sonnet and Haiku, when to adjust effort instead of switching models, and how to cut costs by running subagents on smaller models.

Three decision areas covered in the guide

Each area is a choice a team makes when designing an AI application

Step 1

Choose the model

When to use each named model: Fable 5.2, Opus 5.5, Sonnet and Haiku.

→

Step 2

Adjust effort first

When to change the effort setting instead of moving to a different model.

→

Step 3

Route subagents smaller

Assign subagents in multi-agent workflows to smaller models to reduce costs.

The model lineup as presented

Shown as a tiered lineup, in the order named. The announcement does not rank capability, speed or cost.

Fable 5.2

Opus 5.5

Sonnet

Haiku

What the announcement leaves out

The guide's detailed criteria are not in the summary. Readers need the full document to act on it.

Selection criteria

Not detailed in the post

Benchmark results

Not included

Pricing figures

Not included

Continue reading

The rest of this article is for AI News Blitz readers. Choose an option below to keep reading.