Reference for selecting the right Claude model (Opus 4.6, Sonnet 4.6, Haiku 4.5) based on task complexity, reasoning depth, and cost/speed tradeoffs. Auto-loaded when model selection is relevant.
Recommend the most appropriate Claude model for each task based on complexity, reasoning requirements, and cost/speed tradeoffs.
| Model ID | Name | Strengths |
|---|---|---|
claude-opus-4-6 | Claude Opus 4.6 | Deep reasoning, complex decisions, nuanced judgment |
claude-sonnet-4-6 | Claude Sonnet 4.6 | Balanced capability and speed, code generation, most tasks |
claude-haiku-4-5-20251001 | Claude Haiku 4.5 | Fast, lightweight, simple and conversational tasks |
Is the task conversational, templated, or config-only?
→ Haiku
Does the task involve code generation, analysis, refactoring, or standard TDD?
→ Sonnet
Does the task require ANY of the following?
- Architectural tradeoffs or design decisions
- Multi-step reasoning across large context
- Nuanced judgment with significant consequences
→ Opus
When in doubt: Sonnet is the safe default.
| Persona | Default Model | Reason |
|---|---|---|
| Product Visionary (Lisa) | Haiku 4.5 | Conversational, exploratory |
| Architect (Professor Frink) | Opus 4.6 | Complex tradeoff analysis |
| Senior Developer / TDD (Sideshow Bob) | Sonnet 4.6 | Code generation + reasoning |
| Senior Tester (Martin Prince) | Sonnet 4.6 | Code analysis and test generation |
/model claude-opus-4-6
/model claude-sonnet-4-6
/model claude-haiku-4-5-20251001