Claude Sonnet 5 Hidden Price Hike: Per-Task Token Consumption Up 40%, Real Cost Doubles
Decision Brief
What changedClaude Sonnet 5 surpasses Opus 4.8 on some tasks, but its per-task token consumption increased by ~40%, nearly doubling actual costs—continuing Anthropic's pattern of hidden price increases.
Why it mattersAI builders must watch for hidden costs from unchanged list prices but higher token usage, to avoid budget overruns.
Who should careTeams building on model APIs
Affected stackClaude
Builder actionMonitor
Source confidenceMedium · Reliable media or first-hand reporting
Claude Sonnet 5 ranks fifth on the Artificial Analysis Intelligence Index with a score of 53, even beating the more expensive Opus 4.8 on some agent-based tasks. However, the model consumes about 40% more tokens per task than its predecessor. Although the list price remains unchanged, the actual cost nearly doubles. This continues Anthropic's pattern of hidden price increases, reminding AI builders to consider token efficiency along with list prices when evaluating models.
Summary basis: official / RSS sourceUnless it says 'full article read', this summary is based only on publicly available content — it never pretends to have read restricted originals.
Sources
- The Decoder：AI News
- The Decoder：AI News