Claude Sonnet pricing: Sonnet 5.5, Sonnet 5 and Sonnet 4.6
Claude Sonnet is the middle of Anthropic's lineup: fast, capable and a fraction of the price of Opus. This page compares the current Sonnet versions, shows what a request costs and explains where Sonnet fits against Opus and Haiku. Part of our AI costs and pricing section.
Prices last verified against official provider pages on . How we verify prices. Written by the Chelure Enterprises Editorial Team.
Prices per million tokens
| Model | Input | Output | Cached input (read) | Cache write (5 min) | Batch API | Announced change |
|---|---|---|---|---|---|---|
| Claude Sonnet 5.5 | $2.00 | $10.00 | $0.10 | $2.50 | Yes (50% off) | - |
| Claude Sonnet 5 | $2.00 | $10.00 | $0.20 | $2.50 | Yes (50% off) | - |
| Claude Sonnet 4.6 | $3.00 | $15.00 | $0.30 | $3.75 | Yes (50% off) | - |
Verified November 13, 2026 against the provider's pricing page. Compare every provider in the price comparison, or model your own workload in the cost calculator.
What each model is for
| Model | Positioning (provider) | Context window | Max output |
|---|---|---|---|
| Claude Sonnet 5.5 | The best combination of speed and intelligence (Anthropic); fast latency | 1M tokens | 128K tokens |
| Claude Sonnet 5 | Previous Sonnet generation | 1M tokens | 128K tokens |
| Claude Sonnet 4.6 | Earlier Sonnet generation | 1M tokens | 128K tokens |
What a request costs
Costs per single request, from the prices above. Agent step uses cache reads at each model's cached-input rate.
| Model | Chat turn (1,500 input / 400 output tokens, no caching) | RAG answer (8,000 input / 500 output tokens, no caching) | Agent step (30,000 input / 2,000 output tokens, 80% of input read from cache) | 1,000 chat turns |
|---|---|---|---|---|
| Claude Sonnet 5.5 | $0.0070 | $0.021 | $0.034 | $7.00 |
| Claude Sonnet 5 | $0.0070 | $0.021 | $0.037 | $7.00 |
| Claude Sonnet 4.6 | $0.011 | $0.032 | $0.055 | $10.50 |
What changed between Sonnet versions
Sonnet 5 and 5.5 list at $2 input and $10 output per million tokens, a third cheaper than Sonnet 4.6 at $3 and $15. The difference between 5 and 5.5 is in caching: Sonnet 5.5 cache reads cost 5 percent of input ($0.10 per million) against 10 percent ($0.20) on Sonnet 5, so a workload with a large cached prefix is cheaper on 5.5 at the same list price.
At $2 and $10 a typical chat turn of 1,500 input and 400 output tokens costs a fraction of a cent, and the table above shows how that scales. For classification-style work at very high volume, compare Haiku; for the hardest agentic tasks, compare Opus. Our GPT vs Claude cost comparison puts Sonnet next to OpenAI's equivalent tier.
Frequently asked questions
How much does Claude Sonnet 5.5 cost?
$2 per million input tokens and $10 per million output tokens. Cache reads are 5 percent of the input price ($0.10 per million), cache writes for the 5-minute cache are $2.50 per million, and Batch API requests are 50 percent off.
What is Sonnet best at?
Anthropic describes Sonnet 5.5 as the best combination of speed and intelligence, with fast comparative latency, behind only Haiku among the current models.
Is Sonnet 5.5 cheaper than Sonnet 4.6?
Yes. Sonnet 4.6 lists at $3 input and $15 output, while Sonnet 5 and 5.5 list at $2 and $10, per Anthropic's models overview.
What is the Sonnet 5.5 context window?
1M tokens with a maximum output of 128K tokens. Sonnet 4.5 had 200K context and 64K output.
Sonnet or Haiku?
Haiku 5.5 is designed for high-volume, latency-sensitive tasks such as classification, extraction and routing, and lists from $0.10 input and $0.50 output per million tokens. Use Sonnet when the task needs more capability than Haiku gives.
Sources
Primary sources opened on November 13, 2026. Prices come from our verified price table (see the methodology).