Claude Sonnet 5 — API Pricing & Benchmarks
Claude Sonnet 5 is Anthropic's mid reasoning model, released in 2026-06. The API costs $2.00 per 1M input tokens and $10.0 per 1M output tokens ($4.00 blended at a 3:1 ratio), with cached input at $0.20. It supports a 1M-token context window. On the independently measured Artificial Analysis Intelligence Index it scores 38.2, ranking #26 of 72 models tracked on this site, with a median output speed of 85 tokens/sec.
Explore provider routes and additional benchmark sources for Claude Sonnet 5
| Provider | Anthropic |
| Input price / 1M tokens | $2.00 |
| Output price / 1M tokens | $10.0 |
| Cached input / 1M tokens | $0.20 |
| Blended price (3:1) | $4.00 |
| Context window | 1M |
| Max output tokens | 128K |
| Knowledge cutoff | 2026-01 |
| License | proprietary |
| Input modalities | T+I |
| Output speed (tok/s) | 85 |
| Released | 2026-06 |
| Artificial Analysis Intelligence Index | 38.2 |
| LMArena Elo | 1442 |
| GPQA Diamond | 91.1% |
| SWE-bench Verified | 85.2% |
| Humanity's Last Exam | 41.3% |
Where Claude Sonnet 5 fits
On a blended 3:1 basis Claude Sonnet 5 costs $4.00 per 1M tokens, cheaper than 26% of the 84 priced models in this index. It sits in the mid-tier bracket where most production chat, RAG and summarization workloads run: meaningfully cheaper than flagships while keeping most of their capability. Cached input is priced at $0.20 (90% below standard input), which rewards prompt structures with long stable prefixes — see our caching guide.
Price history
| Period | Input $/1M | Output $/1M | Cached $/1M |
|---|---|---|---|
| 2026-07-18 → current | $2.00 | $10.0 | $0.20 |
No list-price changes recorded since tracking began. Full dataset: price-history.json (CC BY 4.0).
Alternatives at a similar price
- GPT-6 Sol (OpenAI) — $4.00/1M blended, AA Index 47.5
- Qwen3.7-Max (Alibaba Qwen) — $3.75/1M blended, AA Index 29.5
- GPT-4o (OpenAI) — $4.38/1M blended, AA Index 9
- Command A (Cohere) — $4.38/1M blended, AA Index 7
About Anthropic
Anthropic builds the Claude family, organized into Opus (flagship), Sonnet (mid-tier) and Haiku (small) tiers alongside the Fable line. Claude models currently lead the Artificial Analysis Intelligence Index and are particularly strong on agentic-coding benchmarks such as SWE-bench Verified.