Claude Sonnet 4.6 Cost Calculator

The production-default Claude model. $3 input / $15 output per 1M tokens with a 1M context window.

Claude Sonnet 4.6 hits the sweet spot for most production deployments: coding agents, customer chat, JSON extraction, and RAG. At $3 / $15 per million tokens it costs 40% less than Opus 4.7 while matching it on most non-reasoning benchmarks.

Combined with prompt caching (cached reads at $0.30/M) and batch pricing (50% off), Sonnet 4.6 is the default Claude model for teams scaling beyond a proof-of-concept.

Models included

Claude Opus 4.7 (Anthropic) — $5.00 input / $25.00 output per 1M tokens · 1M context window
Claude Sonnet 4.6 (Anthropic) — $3.00 input / $15.00 output per 1M tokens · 1M context window
Claude Haiku 4.5 (Anthropic) — $1.00 input / $5.00 output per 1M tokens · 200K context window

Frequently asked questions

How much does Claude Sonnet 4.6 cost?

Claude Sonnet 4.6 costs $3 per million input tokens and $15 per million output tokens. Cached input reads are $0.30 per 1M.

Is Sonnet 4.6 cheaper than GPT-5.5?

Yes. Sonnet 4.6 is $3/$15 per 1M vs GPT-5.5 at $5/$30 — roughly 40% cheaper across the board. Quality is comparable on most coding and chat benchmarks.

What is the context window for Sonnet 4.6?

Claude Sonnet 4.6 supports a 1-million-token context window at standard pricing.

When should I upgrade from Sonnet 4.6 to Opus 4.7?

Move to Opus only when you measurably need its reasoning advantage — long multi-step research, deep code refactors, or hard math/logic. For 90% of production work, Sonnet ships the same answers at 40% less.