8 models tracked
Anthropic API pricing
Anthropic prices simply and documents thoroughly, but publishes no browser-capable tokenizer — so token counts taken before you send are estimates unless you call their server-side endpoint.
- Models tracked
- 8 7 current
- Cheapest input
- $1.00 Claude Haiku 4.5
- Highest input
- $10.00 Claude Fable 5
- Exact token counts
- 0 / 8 no browser tokenizer
Every Anthropic model, cheapest first
| Model | Input | Cached | Output | Context | Count | Chat turn |
|---|---|---|---|---|---|---|
| Claude Haiku 4.5 | $1.00 | $0.1000 | $5.00 | 200K | Estimate | $0.003000 |
| Claude Sonnet 5 | $2.00 | $0.2000 | $10.00 | 1M | Estimate | $0.006000 |
| Claude Sonnet 4.6 | $3.00 | $0.3000 | $15.00 | 1M | Estimate | $0.009000 |
| Claude Sonnet 4.5 legacy | $3.00 | $0.3000 | $15.00 | 200K | Estimate | $0.009000 |
| Claude Opus 5 | $5.00 | $0.5000 | $25.00 | 1M | Estimate | $0.0150 |
| Claude Opus 4.8 | $5.00 | $0.5000 | $25.00 | 1M | Estimate | $0.0150 |
| Claude Opus 4.6 | $5.00 | $0.5000 | $25.00 | 1M | Estimate | $0.0150 |
| Claude Fable 5 | $10.00 | $1.00 | $50.00 | 1M | Estimate | $0.0300 |
Chat turn is 1,500 input and 300 output tokens. Prices read from Anthropic’s pricing documentation; each row links to its own dated source.
What is distinctive about Anthropic pricing
- 01
Prompt caching uses explicit multipliers: 1.25× base for a five-minute cache write, 2× for a one-hour write, and 0.1× for a read. One cache hit already puts you ahead on the short-lived option.
- 02
Batch processing is discounted 50% on both input and output, and stacks with caching.
- 03
Long context is included at standard pricing on models with a one-million token window — a 900,000 token request bills at the same per-token rate as a 9,000 token one.
Token counting on Anthropic
No public browser tokenizer exists. Anthropic has stated that the tokenizer introduced with Claude 4.7 produces approximately 30% more tokens for the same text than its predecessor, which means a per-page cost comparison across that boundary needs re-measuring rather than re-pricing.
Counts shown here for Anthropic models are estimates and are labelled as such everywhere they appear. The methodology page documents every scaling factor used.
Compare with other providers
Or shortlist across all of them at once in the model finder, filtering by budget, context window and whether token counts can be exact.
Frequently asked questions
- How much does the Anthropic API cost?
- It depends on the model. Across the 8 models tracked here, input ranges from $1.00 to $10.00 per million tokens and output from $5.00 to $50.00. On a typical chat turn of 1,500 input and 300 output tokens the range is $0.003000 to $0.0300 per request.
- What is the cheapest Anthropic model?
- Claude Haiku 4.5, at $1.00 per million input tokens and $5.00 per million output. Whether it is cheap enough depends on whether it can do your task — the cost of finding out is usually a few cents.
- Can I count Anthropic tokens exactly?
- Not in a browser. No public browser tokenizer exists. Anthropic has stated that the tokenizer introduced with Claude 4.7 produces approximately 30% more tokens for the same text than its predecessor, which means a per-page cost comparison across that boundary needs re-measuring rather than re-pricing. Counts shown on this site for Anthropic models are labelled as estimates everywhere they appear.
- How current is this Anthropic pricing?
- Every model entry stores the URL it was read from and the date it was read, both shown in the table above. The full data set was last reviewed on August 5, 2026. Prices change without notice, so verify against Anthropic's own documentation before committing spend.