Claude Opus 4.1 vs Opus 5.5: price per token and per task
Short answer: at list prices Claude Opus 4 and 4.1 cost $15 per million input tokens and $75 per million output tokens. Claude Opus 5.5 costs $4 and $20, 3.75 times less per token. Opus 5.5 uses a newer tokenizer that counts about 30% more tokens for the same text, so for the same work the list-price gap is closer to 2.9 times.
| Model | Input | Output | Cache read | Cache write 5m / 1h | Tokenizer |
|---|---|---|---|---|---|
| Claude Opus 4 / 4.1 | $15 | $75 | $1.50 | $18.75 / $30 | previous |
| Claude Opus 4.5 / 4.6 | $5 | $25 | $0.50 | $6.25 / $10 | previous |
| Claude Opus 4.7 / 4.8 / 5 | $5 | $25 | $0.50 | $6.25 / $10 | newer |
| Claude Opus 5.5 | $4 | $20 | $0.20 | $5 / $8 | newer |
USD per million tokens, standard tier, global routing.
Price per token is not price per task. Anthropic notes that Claude 4.7 and later models use a newer tokenizer that produces about 30% more tokens for the same text, depending on the content (pricing page). Output length and quality on your own tasks also change the result, so treat the numbers below as a scenario and test on your own prompts.
A worked scenario
Suppose a team spent $3,000 last month on Opus 4.1. The same work at Opus 5.5 list prices, with 30% more tokens, would be about $3,000 × 4/15 × 1.3 ≈ $1,040. That is a scenario, not a saving already made: output length, cache hit rate and quality on your tasks can move it either way.
The same arithmetic shows why moving from Opus 4.5 or 4.6 to Opus 5.5 may not save anything: the per-token price drops by 20%, while the newer tokenizer can add about 30% more tokens.
What to do before switching
- Take a sample of real prompts and run them on both models.
- Compare quality, output length and token counts, not only the per-token price.
- Change the model name in one service first and watch the bill for a week.
Check your own bill
Export last month's usage or cost CSV from the Claude Console and open it in the free checkup. It shows how much went to each model and a price scenario for older versions, with the method next to the number. The file stays on your device.
Questions
How much more does Claude Opus 4.1 cost than Opus 5.5?
Per token, 3.75 times more at list price: $15 versus $4 per million input tokens and $75 versus $20 per million output tokens. Because Opus 5.5 uses a newer tokenizer that produces about 30% more tokens for the same text, the gap for the same work is closer to 2.9 times at list prices.
Is Opus 5.5 the same tier as Opus 4.1?
Both are in the Opus line, Anthropic's top tier, and Opus 5.5 is the current version. Test quality and token counts on your own tasks before switching.
How do I see how much I spend on Opus 4.1?
Export the usage or cost CSV from the Claude Console and open it in the free checkup at aispenddoctor.com. It shows spend per model and a price scenario for older versions.
List prices checked 27 September 2026 on the official Anthropic and OpenAI price pages. Prices change; check the source before making decisions. Not affiliated with Anthropic or OpenAI. © 2026 Valeriy Danilov. AI Spend Doctor by AgentBridge Labs.