★ OverviewIntermediate10 min
The Claude Model Family — When Each Model Excels
Opus 4.7, Sonnet 4.6, Haiku 4.5 — three capability tiers with real cost differences. Understanding when each model earns its price tag is one of the highest-leverage decisions in your Claude Code workflow.
Quick Reference
- →Opus 4.7: complex reasoning, architecture, security — use when first-answer quality is load-bearing
- →Sonnet 4.6: everyday coding, code review, feature implementation with clear patterns
- →Haiku 4.5: ~19x cheaper than Opus — exploration, quick lookups, tasks you will verify anyway
- →Aliases: opus, sonnet, haiku — always point to latest version; avoid pinning specific model IDs
- →opusplan alias: Opus for planning phase, Sonnet for execution — optimal cost/quality split
- →Default (April 2026): Enterprise/API users default to Opus 4.7, override with ANTHROPIC_MODEL
- →/model command: switch models mid-session without losing context
The Three Tiers
| Model | Released | Context | Relative cost | Best for |
|---|---|---|---|---|
| Opus 4.7 | April 16, 2026 | 200K | Highest | Complex reasoning, architecture, security review |
| Sonnet 4.6 | Feb 17, 2026 | 200K | Mid | Everyday coding, code review, feature implementation |
| Haiku 4.5 | 2025 | 200K | ~19x less than Opus | Exploration, quick questions, mechanical tasks |
Opus 4.7 was described at release as a step-change improvement in agentic coding. As of April 23, 2026, it became the default for Enterprise pay-as-you-go and Anthropic API users. Sonnet 4.6 remains the default on Pro, Max, and Team plans — it handles most everyday development tasks without the Opus cost premium.