Model Details
Where is Claude Sonnet 5.5 a perfect fit?
- Software engineering feature development, debugging, refactoring, and code review
- Coding agents — long-running, well-scoped autonomous development
- Production agents — investigation, review, drafting, and other repeatable workflows
- Computer use — interacting with applications and graphical interfaces
- Design work — polished UI, diagrams, user flows, and visual artifacts
- Knowledge work, analysis, research, extraction, and forecasting
- Documents & presentations — polished reports, slides, and spreadsheets
- High-volume workloads, applications where speed and token efficiency matter
Claude Sonnet 5.5 is positioned as the faster, lower-cost complement to Opus 5.5
Quick Model Estimate
Your Claude Sonnet 5.5 Cost Estimate
💰 Total Cost
—
for 1000 input + 1000 output tokens
Cost Breakdown
Prices are watched for changes and checked against the provider's own page.
Pricing
|
Provider
↕
|
Modality
↕
|
Service Tier
↕
|
Input Price
(per 1M tokens) ↕ |
Output Price
(per 1M tokens) ↕ |
Cached Input
(per 1M tokens) ↕ |
Cached Output
(per 1M tokens) ↕ |
Context Size
↕
|
View
|
|---|---|---|---|---|---|---|---|---|
 Anthropic | Text | Standard | $2.0000 | $10.0000 | $0.1000 | $2.5000 | 1,000,000 tokens | → |
| Anthropic | Text | Batch | $1.0000 | $5.0000 | $0.0500 | $1.2500 | 1,000,000 tokens | → |
 Perplexity | Text | Standard | $2.0000 | $10.0000 | $0.1000 | N/A | 1,000,000 tokens | → |
Price changes
Every recorded change to Claude Sonnet 5.5’s price, most recent first. Changes across all models are on the changes page.
-
Perplexity cut the price of Claude Sonnet 5.5, effective 13 October 2026.
Cached input $0.20 → $0.10 per 1M tokens (-50%).
Per-configuration detail on Perplexity → -
Anthropic cut the price of Claude Sonnet 5.5, effective 8 October 2026.
Cached input $0.20 → $0.10 per 1M tokens (-50%). 2 configurations changed.
Per-configuration detail on Anthropic →
Benchmarks
Scores from standardized evaluations by Artificial Analysis and Design Arena. Higher is better — the indices summarize overall ability, while the detailed scores break down performance on individual benchmarks.
Artificial Analysis
Higher is better · benchmarked by Artificial AnalysisExpert-level questions across many academic domains
Unpublished physics research reasoning problems
Long-context reasoning across large inputs
Research-level scientific coding tasks
Economically valuable, real-world knowledge work
Breadth of factual knowledge across domains
How reliably the model avoids fabricated answers
| Variant | Humanity's Last Exam | AA-LCR | SciCode | CritPt | GDPval | Omniscience Accuracy | Omniscience Non-Hallucination |
|---|---|---|---|---|---|---|---|
| Claude Sonnet 5.5 (Max, Default Fallback) | 55% | 82.7% | 61% | 31.4% | 66.9% | 53.9% | 53% |
| Claude Sonnet 5.5 (Xhigh, Default Fallback) | 50% | 79.7% | 57.3% | 31.1% | 61.4% | 53% | 37.1% |
| Claude Sonnet 5.5 (High, Default Fallback) | 45.8% | 78% | 53.7% | 24.6% | 52.4% | 52% | 35.4% |
| Claude Sonnet 5.5 (Medium, Default Fallback) | 39.8% | 76.3% | 52.9% | 16.9% | 41.1% | 47.2% | 48.8% |
| Claude Sonnet 5.5 (Low, Default Fallback) | 36.2% | 76% | 49.1% | 11.4% | 33.9% | 46.3% | 49.8% |
Design Arena
Elo rating by arena & category| Variant | Arena | Category | Elo | Win % | Percentile | Avg time (ms) |
|---|---|---|---|---|---|---|
| claude-sonnet-5-5 | agents | blender3d | 1,300 | 63.6% | 88 | — |
| claude-sonnet-5-5 | agents | mobileapps | 1,286 | 60% | 94 | — |
| claude-sonnet-5-5 | agents | spreadsheet | 1,291 | 53.3% | 84 | — |
| claude-sonnet-5-5 | agents | webapps | 1,305 | 60.3% | 91 | — |
| claude-sonnet-5-5 | models | 3d | 1,399 | 51.3% | 97 | 576,418 |
| claude-sonnet-5-5 | models | codecategories | 1,354 | 56.4% | 96 | 391,124 |
| claude-sonnet-5-5 | models | dataviz | 1,370 | 65.7% | 99 | 248,776 |
| claude-sonnet-5-5 | models | gamedev | 1,414 | 62.3% | 99 | 479,970 |
| claude-sonnet-5-5 | models | uicomponent | 1,373 | 63.1% | 97 | 227,671 |
| claude-sonnet-5-5 | models | website | 1,323 | 54.7% | 96 | 282,235 |
Compare Claude Sonnet 5.5 with
Two models side by side: worked costs, every rate, and every benchmark figure both of them report.
Other Models in the Claude 5.5 Family
FAQs about Claude Sonnet 5.5
How much does Claude Sonnet 5.5 cost per 1M tokens?
Claude Sonnet 5.5 costs between $1.00 and $2.00 per million input tokens, and between $5.00 and $10.00 per million output tokens. The range covers Standard and Batch pricing.
What does a typical workload cost with Claude Sonnet 5.5?
1,000 requests of 2,000 input and 500 output tokens each — 2,000,000 input and 500,000 output tokens in total — costs $4.50 with Claude Sonnet 5.5 at its lowest rates. Use the calculator on this page for your own volumes.
Has Claude Sonnet 5.5's price changed?
Yes, on 2 occasions. Most recently Perplexity cut it on 13 October 2026, from $0.20 to $0.10 per million cached input tokens.
Which provider is cheapest for Claude Sonnet 5.5?
Anthropic has the lowest input rate for Claude Sonnet 5.5, at $1.00 per million tokens. Across all providers: Anthropic from $1.00 and Perplexity from $2.00 per million input tokens.
Is there cheaper Claude Sonnet 5.5 pricing than the standard rate?
Yes. Batch pricing for Claude Sonnet 5.5 starts at $1.00 per million input tokens against $2.00 on the standard tier, about 50% lower.
What is Claude Sonnet 5.5's context window?
Claude Sonnet 5.5 has a context window of 1,000,000 tokens, and returns up to 128,000 tokens in a single response.
What input types does Claude Sonnet 5.5 support?
Claude Sonnet 5.5 accepts text, image and file input, and returns text.