Claude Haiku 5.5 has been released as the new default AI model for everyone, replacing Haiku 4.5. The change brings consistent model behavior across all AI surfaces including AI Task Writer, Agent Chat, and code-generation workflows.
Model choice and behavior:
- Haiku 5.5 now serves as the default model for all new AI conversations
- The replacement reduces confusion around having multiple Haiku versions live simultaneously
- Model quality reflects real-world usage patterns and usage-based cost
Real-world cost: In live measurements on the QA Sandbox board, Claude Haiku 5.5 costs approximately 1.7x more per request than GPT 6 Luna. This difference stems from longer answer lengths rather than per-token pricing parity:
AI chat, 10 prompts each on the same board:- Haiku 5.5: 1.70x cost of GPT 6 Luna per request- All models use identical per-token pricingThe model is fully operational on all supported environments including Free and paid subscription tiers.
When Haiku 5.5 is selected (or left as default), it behaves equivalently to GPT 6 Luna in all supported AI workflows, ensuring predictable outputs regardless of model provider.
https://app.hypertask.ai/detail/project-15/7010