Claude Sonnet 5
by Anthropic · anthropic/claude-sonnet-5 · #347 cheapest of 422 paid models
Prices updated Sep 20, 2026, 12:34 AM UTC · refreshed hourly
Sonnet 5 is Anthropic's most capable Sonnet-class model, with frontier performance across coding, agents, and professional work. It supports adaptive thinking with selectable reasoning effort levels (low, medium, high, max,...
Compare vs:GPT-5.6 SolGPT-5.6 TerraGPT-4.1GPT-4.1 Mini
Pricing
Input
$2.00
per 1M tokens
Output
$10.00
per 1M tokens
Blended (3:1)
$4.00
3 input : 1 output
Cache read
$0.20
per 1M cached tokens
Cache write
$2.50
per 1M tokens
Cache & batch economics
Effective prices for Claude Sonnet 5 with prompt caching.
| Cache hit rate | Effective blended $/1M | Example request* | vs no cache |
|---|---|---|---|
| 0% | $4.00 | $0.214 | — |
| 50% | $3.33 | $0.124 | −42% |
| 90% | $2.79 | $0.052 | −76% |
*Example: 100K sticky context + 2K new input + 1K output. Batch discount is an approximate 50% off list rates for this provider’s batch API. Confirm on the vendor’s pricing page.
What a request costs
| Workload | Input tokens | Output tokens | Cost / request | Cost / 1K requests |
|---|---|---|---|---|
| Short chat message | 500 | 300 | $0.004 | $4.00 |
| Document summary | 8,000 | 1,000 | $0.026 | $26.00 |
| Codebase question (RAG) | 30,000 | 2,000 | $0.08 | $80.00 |
| Long-context analysis | 150,000 | 5,000 | $0.35 | $350.00 |
Performance
Intelligence Index
38.2
AA composite quality
Coding Index
71.5
Math Index
—
Agentic Index
44.3
Output speed
79 tok/s
median
Time to first token
116.86s
median
Time to first answer
116.86s
after reasoning tokens
Artificial Analysis benchmarks
| Benchmark | Score |
|---|---|
| GPQA Diamond | 91.1% |
| Humanity's Last Exam | 41.3% |
| SciCode | 54.3% |
| AA-LCR (Long Context Reasoning) | 82.0% |
| Terminal-Bench 2.1 | 80.5% |
| τ³-Bench Banking | 37.3% |
Design Arena Elo
| Category | Elo |
|---|---|
| Website | 1,289 |
| Web apps | 1,260 |
| Full-stack | 1,260 |
| Mobile apps | 1,250 |
| Android native | 1,241 |
| UI components | 1,298 |
| Data viz | 1,260 |
| SVG | 1,212 |
| 3D | 1,290 |
| Game dev | 1,313 |
| Agentic game dev | 1,226 |
| Godot game dev | 1,238 |
| HTML slides | 1,220 |
| Python→PPTX slides | 1,216 |
| Code categories | 1,293 |
| ASCII art | 1,226 |
Specs
Context window
1M
tokens
Max output
128K
tokens
Input modalities
text, image, file
Tokenizer
Claude