Claude Sonnet 5 pricing
Claude Sonnet 5 costs $2 per million input tokens and $10 per million output tokens , with cached input at $0.2. Its context window is 1M tokens. Batch requests are discounted 50%.
Price per million tokens
| Input | $2 |
| Cached input | $0.2 |
| Output | $10 |
| Context window | 1M tokens |
| Batch discount | −50% |
| Minimum prefix to cache | 1,024 tokens |
| Tokenizer factor | ×1.3 |
The $2/$10 introductory rate is now standard; the planned 1 Sep rise was cancelled
What it costs in practice
Monthly cost for three common workload shapes, with prompt caching applied.
| Workload | Shape | Per request | Per month |
|---|---|---|---|
| Support chatbot | 1.5k in / 400 out, 100k requests/month, 30% cached | $0.00805 | $804.70 |
| RAG / document Q&A | 12k in / 700 out, 50k requests/month, 60% cached | $0.0235 | $1,173 |
| Coding agent | 25k in / 2.5k out, 20k requests/month, 80% cached | $0.0507 | $1,014 |
Model your own workload with caching, batching and tokenizer correction →
Caching on Claude Sonnet 5
A cached prefix on this model must be at least 1,024 tokens. Below that, Anthropic ignores cache_control silently — no error is returned and you pay full input price on
every request. A cache hit costs $0.2 per million tokens, roughly 10% of the
base input rate.
Comparable models from other providers
| Model | Provider | Input | Output |
|---|---|---|---|
| GPT-5.6 Terra | OpenAI | $2 | $12 |
| GPT-4.1 | OpenAI | $2 | $8 |
| o3 | OpenAI | $2 | $8 |
| Gemini 3.1 Pro (preview) | $2 | $12 |
All figures verified 2026-08-18 against Anthropic's official pricing documentation . Fields the provider does not publish are shown as “—” rather than estimated.