Claude Fable 5 Pricing: $10/$50 per Million Tokens, Less Than Half of Mythos Preview
Anthropic priced Claude Fable 5 at $10 per million input tokens and $50 per million output tokens - undercutting Claude Mythos Preview by more than half while doubling Opus 4.8.
Anthropic has priced its most powerful public model aggressively. Claude Fable 5 launches at $10 per million input tokens and $50 per million output tokens - less than half the price of Claude Mythos Preview, the restricted model it effectively succeeds in public-facing form.
Where Fable 5 Sits in the Lineup
The new price point places Fable 5 at exactly double Claude Opus 4.8, which costs $5 per million input tokens and $25 per million output tokens.
| Model | Input / 1M tokens | Output / 1M tokens |
| Claude Fable 5 | $10 | $50 |
| Claude Opus 4.8 | $5 | $25 |
| Claude Mythos Preview | More than 2x Fable 5 | More than 2x Fable 5 |
For buyers, the calculus is straightforward: Fable 5 scores 80.3% on SWE-Bench Pro versus Opus 4.8's 69.2% - roughly 11 points - for twice the per-token cost. For teams whose workloads involve long, complex tasks, where Anthropic says Fable 5's advantages are largest, the premium may pay for itself in fewer retries and fewer human interventions. Equinox CTO Luke Anderson made exactly that argument: "Claude Fable 5 delivers more capable engineering in fewer turns than prior models."
A Free On-Ramp
Anthropic is softening the launch with a free access window. From June 9 through June 22, 2026, Claude Fable 5 is included at no extra cost on Pro, Max, Team, and Enterprise plans. From June 23, using the model requires usage credits. The model is fully available on the Claude API from day one, and generally available on Amazon Bedrock and GitHub Copilot.
The Mythos Preview Comparison
The most notable line in the pricing announcement is the comparison to Claude Mythos Preview. Fable 5 - the same underlying model as Claude Mythos 5, made safe for general use - costs less than half as much as the preview-era Mythos offering. That is a significant repricing of frontier capability: what was recently a restricted, premium product is now a generally available model at a mid-tier frontier price.
The signal to the market is clear. Anthropic believes it can win on capability-per-dollar at the top end, and it is betting that the 11-point benchmark lead, combined with halved Mythos-class pricing, will pull high-value agentic workloads onto the Claude API.