Claude Fable 5 vs Sonnet 5

Claude Fable 5 and Claude Sonnet 5 are both Anthropic models, but they serve different purposes. Anthropic has since released Claude Fable 5.1 and Claude Sonnet 5.5, which list the same base token prices as the two models compared here. Fable 5 is Anthropic's most capable public model for demanding knowledge work and coding. Sonnet 5, released June 30, 2026, is a mid-tier model designed for high-volume agentic tasks and positioned between Haiku and Opus.

So the real question isn't "Which model is better?" It's "Which model fits the task?" Sonnet 5 is the default on Anthropic's Free and Pro plans and is designed to handle many agent steps at a lower cost. Fable 5 makes more sense when the task is difficult enough to justify the higher price per token.

Claude Fable 5 costs $10 per million input tokens and $50 per million output tokens. Claude Sonnet 5 costs $2 per million input tokens and $10 per million output tokens at its standard rate. Anthropic withdrew the increase to $3 for input and $15 for output planned for September 1, 2026. That means Sonnet 5 costs one fifth as much as Fable 5.

This page breaks down what each model is designed to do, how their prices compare, and a simple way to choose between them, so you don't overpay for routine work or choose too little power for a difficult task.

Claude Fable 5 vs. Claude Sonnet 5: Side-by-Side

Are you one of these vendors? Update your listing

What Each Model Is Built For

Claude Fable 5 is Anthropic's top-tier model, built for the hardest coding, refactor, and knowledge-work problems, where a wrong step is expensive and the extra per-token cost is worth paying for the strongest available model.

Claude Sonnet 5 is a different bet: a mid-tier model tuned to be Anthropic's "most agentic Sonnet" — able to plan, use tools, and push through long multi-step jobs at a fraction of Fable 5's price. It is also the default model most Claude users are already on, since it powers the Free and Pro plans.

Same vendor, same platform, different economics. Fable 5 is priced for occasional, high-stakes use; Sonnet 5 is priced for volume.

• Claude Fable 5: the flagship, for your hardest problems
• Claude Sonnet 5: the agentic workhorse, for everyday and high-volume tasks

Fable 5.1 and Sonnet 5.5: What Changed

Compare Claude Fable 5.1 and Claude Sonnet 5.5 if you are deploying today, because base rates remain unchanged while cache reads cost less. Claude Fable 5 costs $10 per million input tokens and $50 per million output tokens, with cache reads at $1 per million. Under Claude Fable 5.1 explained, the newest Fable keeps those base rates while cache reads fall to $0.25 per million, or 0.025 times input.

Claude Sonnet 5 costs $2 per million input tokens and $10 per million output tokens with $0.20 cache reads, following the withdrawal of a planned increase to $3 and $15. For the newest Sonnet model, Anthropic cut cache reads to $0.10 per million tokens on October 7, 2026, while preserving the $2 and $10 base rates. Read Claude Sonnet 5.5 explained for a breakdown of that change.

Sonnet remains one fifth the base price of Fable. Fable 5 suits the hardest coding and knowledge work, while Sonnet suits high-volume agentic work. You can compare both against the entire catalog in Claude models compared.

Price Comparison

Claude Sonnet 5 costs one fifth of Claude Fable 5 on both input and output tokens. At its standard rate, Sonnet 5 is $2 input / $10 output per million tokens, versus Fable 5's $10 / $50.

Anthropic withdrew the increase to $3 / $15 it had planned for September 1, 2026, so that standard rate is the current price. One billing detail to plan for: Sonnet 5 uses a newer tokenizer that produces roughly 30% more tokens for the same text, per Anthropic, so the effective gap is slightly smaller than the headline per-token prices suggest.

For an agent that runs dozens of steps per task, that price gap compounds fast. Running the same workload on Fable 5 instead of Sonnet 5 can mean paying several times more for work that Sonnet 5 handles adequately.

→ Compare per-token pricing for every AI model

When to Choose Claude Fable 5 vs Sonnet 5

Choose Claude Fable 5 when the task is genuinely hard — a complex refactor, a tricky debugging session, or research where a wrong answer is costly, and the model's extra reasoning strength is worth the higher price.

Choose Claude Sonnet 5 for the bulk of everyday work: routine coding, document drafting, summarization, and agent workflows that run many steps. Its lower cost and native agentic design make it the practical default for volume.

Many teams run both: Sonnet 5 as the default for most requests and agent steps, with an escalation path to Fable 5 for the small share of tasks that actually need the flagship model.

• Pick Fable 5: hardest coding and knowledge-work problems, cost is secondary
• Pick Sonnet 5: high-volume agent workloads, everyday business tasks, cost-sensitive
• Use both: Sonnet 5 as the default, escalate to Fable 5 only when a task needs it

How to Use Claude Fable 5 and Claude Sonnet 5

To use Claude Fable 5 and Claude Sonnet 5 effectively, start by choosing an interface that exposes the exact model you need for your workload. Test a new configuration on a single representative task, such as drafting a routine email or parsing a short document, before expanding to full automation. Inspect the saved output directly to confirm formatting, accuracy, and adherence to your prompt instructions.

For coding tasks, Cursor serves as an editor option when the target model appears on its supported list. Check current model availability before setup to confirm the exact model is listed. Once connected, save your prompts and review the generated diffs locally before applying changes to your main codebase.

The Verdict

Claude Fable 5 and Claude Sonnet 5 are not rivals so much as two settings on the same dial. Sonnet 5 is the practical default for volume and everyday agent work; Fable 5 is the model to reach for on the hardest problems where the extra cost is worth it.

Not sure which model your workflows actually need, or whether you are overpaying by defaulting to the flagship for routine work? Layer3 Labs can map your tasks to the right Claude model.