Anthropic Claude models: where each one earns its price
| # | AI model | Capabilities | API $/1M (in / out) | Context window | Price / mo | Output $/1M | Overview | Visit site |
|---|---|---|---|---|---|---|---|---|
| 1 |  Claude Fable 5.1 Anthropic |    | $10 / $50 | 1M tokens | API only | 50 | View | |
| 2 | Claude Opus 5.5 Anthropic | | $4 / $20 | 1M tokens | API only | 20 | View | |
| 3 | Claude Sonnet 5.5 Anthropic | | $2 / $10 | 1M tokens | API only | 10 | View | |
| 4 | Claude Opus 5 Anthropic | | $5 / $25 | 1M tokens | API only | 25 | View | |
| 5 | Claude Fable 5 Anthropic | | $10 / $50 | 1M tokens | API only | 50 | View | |
| 6 | Claude Opus 4.8 Anthropic | | $5 / $25 | 1M tokens | API only | 25 | View | |
| 7 | Claude Sonnet 5 Anthropic | | $2 / $10 | 1M tokens | API only | 10 | View | |
| 8 | Claude Sonnet 4.6 Anthropic | | $3 / $15 | 1M tokens | API only | 15 | View | |
| 9 | Claude Sonnet 4.5 Anthropic | | $3 / $15 | 1M tokens | API only | 15 | View | |
| 10 | Claude Haiku 4.5 Anthropic | | $1 / $5 | 200K tokens | API only | 5 | View | |
| 11 | Claude Opus 4.6 Anthropic | | $5 / $25 | 1M tokens | API only | 25 | View | |
| 12 | Claude Opus 4.7 Anthropic | | $5 / $25 | 1M tokens | API only | 25 | View | |
| 13 | Claude Opus 4.5 Anthropic | | $5 / $25 | 200K tokens | API only | 25 | View |
Browse by need
Anthropic runs a narrow, clearly tiered line-up: Fable 5.1 at the top for long-horizon agentic work, Opus 5 for demanding coding and analysis, Sonnet 5 as the balanced default, Haiku 4.5 for speed. Output spans $5 to $50 per 1M tokens, and the whole current range carries a 1M-token window.
Two economics matter here. Sonnet 5 kept its introductory $2 / $10 pricing after the planned rise was cancelled, making it the strongest value in the mid tier. And cache reads on Fable 5.1 cost 2.5% of base input rather than the usual 10%, which changes the maths for agents with large stable prompts.
Models from this vendor
13 Anthropic models in this rating. Prices are input and output per 1M tokens from the vendor official price list; for dirhams multiply by 3.6725. Tap a card to open the model page.
How to choose
- Pick the tier by consequence: Haiku for volume, Sonnet 5 as the default, Opus 5 for coding and analysis, Fable for long agentic runs.
- Use adjustable thinking rather than switching models — the same Claude at lower depth is often the cheaper answer.
- Structure prompts so the stable part comes first; cache reads cost a tenth of base input, and 2.5% on Fable 5.1.
- Watch the 1.1x multiplier if you pin US inference geography on 4.6-class models.
Common mistakes
Who this suits
Related filters
Frequently asked questions
Which Claude model is best value?
Is there a free Claude API tier?
What is Fable for?
What is the difference between Fable and Opus?
Does Claude have a batch discount?
Frequently searched
Rate this page
Help us improve