Claude Opus 5: what it costs to run
The price Anthropic lists for Claude Opus 5, what a support reply, a document summary and an agent task cost at it, and where that puts it among 37 models.
US dollars per million tokens, standard rate, shortest context band.
- 1,000 support replies
- $25.00
- Cached input
- $0.50
- Context window
- 1,000,000 tokens
- Released
- 24 Jul 2026
- Weights
- Closed
From Anthropic’s own pages, 22 September 2026.
Claude Opus 5 costs $5 per million input tokens and $25 per million output tokens on Anthropic’s own API, and $0.50 for cached input. At those prices 1,000 support replies cost $25.00, the 4th most expensive of 33 priced models on our list.
What a job costs
Three jobs, 1,000 runs of each, at the list price above, and where each puts Claude Opus 5 among the 33 models with a list price.
| Job | 1,000 runs | Among 33 |
|---|---|---|
| Support reply3,000 in, 400 out. | $25.00 | 4th most expensive |
| Document summary25,000 in, 1,000 out. | $150 | 4th most expensive |
| Agent task80,000 in (70,000 cached), 3,000 out. | $160 | 5th most expensive |
The token counts are ours, chosen to look like real work; the prices are Anthropic’s. All three jobs stay under 200,000 tokens a request, below every long-context surcharge on the list.
Claude Opus 5 on real small-business tasks
What the AIGROW benchmark measured when Claude Opus 5 was given 20 office tasks, each marked pass or fail by fixed checks.
- Success rate
- 85% of 20 runs passed (95% interval 64 to 95%), 10th of 36 models.
- Cost per 1,000 passes
- $16.89, failed runs included, worked from list price.
- Answer time
- 27 s median, 83 s at the 90th percentile.
- Strongest and weakest
- Appointment scheduling (100% passed) and quotes (0%).
The small print
What the base rate above leaves out, copied from the maker’s pages: surcharges, discounts, promotions and cache pricing.
Cache write $6.25 (5 min) or $10 (1 h). Fast mode $10 / $50. Batch API 50% off. US-only inference (inference_geo) bills 1.1x. Full context window billed at standard rates.
Claude Opus 5 against the models it is weighed with
Each comparison puts both list prices side by side and costs the same three jobs on each.
- vs Claude Fable 5.1
- Fable costs twice as much as Opus, so buyers ask what the premium buys.
- vs GPT-5.6 Sol
- The two premium workhorses teams pick for agents and long documents, with close list prices.
- vs Gemini 3.1 Pro (preview)
- A common premium choice for writing and analysis, where Gemini costs well under half.
- vs Grok 4.7
- Grok 4.7 is pitched at coding and knowledge work, Opus's core use cases, for far less per token.
- vs DeepSeek V4 Pro
- A low-cost open-weight frontier model against the premium closed model many businesses default to.
- vs Kimi K3
- Kimi K3 is marketed for long-horizon coding and knowledge work, the same ground as Opus.
- vs Claude Sonnet 5
- Tests whether Opus is worth 2.5 times the price of Sonnet on everyday tasks.
Other Anthropic models
| Model | Input | Output | Context | 1,000 replies |
|---|---|---|---|---|
| Claude Haiku 4.5Anthropic | $1.00 | $5.00 | 200K | $5.00 |
| Claude Sonnet 5Anthropic | $2.00 | $10.00 | 1M | $10.00 |
| Claude Sonnet 4.6Anthropic | $3.00 | $15.00 | 1M | $15.00 |
| Claude Fable 5.1Anthropic | $10.00 | $50.00 | 1M | $50.00 |
Where the price comes from
Read on 22 September 2026. Each price is the standard rate for the shortest context band, with no batch or priority discount; long-context surcharges, batch rates and promotions are in each model’s small print. No price was taken from a reseller.
Run Claude Opus 5 for a quarter of its list price
AIGROW API credit covers Claude Opus 5 and is metered at the list price above: $25 buys $100 of usage.
Prices checked 22 September 2026