Claude Opus 5.5 API Pricing & Monthly Cost Simulator
Verified 2026 unit economics for Anthropic Claude Opus 5.5 (anthropic/claude-opus-5.5): $4/1M uncached input, $0.2/1M cached prefix read, and $20/1M output across a 1,000,000-token context window (released 2026-09-22).
Uncached Input (/1M)
$4
Cached Prompt Read (/1M)
$0.2 (-95%)
Output Completion (/1M)
$20
50% Batch API (In / Out)
$2 / $10
Interactive 30-Day Monthly Spend Simulator
Live Unit-Cost & Prompt Cache Slider (Claude Opus 5.5)
anthropic/claude-opus-5.5
Estimated 30-day spend
$732.00/mo
Cost per 1,000 calls
$24.4000
Monthly savings vs 0% cache
$228.00/mo
Technical Limits & API Capabilities
Context window
1,000,000
Max output tokens
128,000
Workload tier
Coding & Agent
Input modalities
text, image, file
Tool Calling: Yes
JSON Schema Strict: Yes
Reasoning Tokens: Supported
50% Batch API: Available
Same-Category 2026 Model Alternatives (Coding & Agent)
| Model | Provider | Input / Cached ($/1M) | Output ($/1M) | Context / Max Out |
|---|---|---|---|---|
| Claude Opus 5.5 | Anthropic | $4 / $0.2 | $20 | 1,000,000 / 128,000 |
| Gemini 4 Hydrogen | $1.5 / $0.075 | $7.5 | 1,048,576 / 943,718 | |
| Gemini 4 Carbon | $2 / $0.1 | $10 | 1,048,576 / 943,718 | |
| Muse Ember | Meta | $0.1 / $0.004 | $0.5 | 1,048,576 / 943,718 |
| GPT-6.1 Terra | OpenAI | $1 / $0.05 | $5 | 1,050,000 / 128,000 |
| MiniMax M3.1 Flash | MiniMax | $0.15 / $0.015 | $0.6 | 1,048,576 / 524,288 |
| Muse Spark Open | Meta | $1.25 / $0.15 | $4.25 | 1,048,576 / 943,718 |
Head-to-Head Matchups Featuring Claude Opus 5.5
- Claude Opus 5.5 vs Claude Sonnet 5.5 — $4 vs $2 / 1M
- Claude Fable 5.2 vs Claude Opus 5.5 — $8 vs $4 / 1M
- GPT-6.1 Terra Pro vs Claude Opus 5.5 — $1 vs $4 / 1M
- Qwen3.8 Max Prime vs Claude Opus 5.5 — $4 vs $4 / 1M
Frequently Asked Questions
How much does Claude Opus 5.5 cost per 1 million tokens in 2026?
Claude Opus 5.5 by Anthropic is billed at $4 per 1M uncached input tokens, $0.2 per 1M cached input read tokens (95% cache discount), and $20 per 1M output tokens.
What are the context window and max output token limits for Claude Opus 5.5?
Claude Opus 5.5 supports a 1,000,000-token input context window and up to 128,000 completion output tokens per request, accepting text, image, file inputs.
Does Claude Opus 5.5 support strict JSON Schema outputs, tool calling, and Batch API?
Claude Opus 5.5 capability profile: tool calling = Supported, strict JSON Schema structured outputs = Supported, reasoning mode = Supported, and 50% asynchronous Batch API = Supported.