Claude Opus 4 8 API pricing

What Claude Opus 4 8 costs in practice

Claude Opus 4 8 from Amazon Bedrock is billed at $5 per million input tokens and $25 per million output tokens. Cached input costs $0.5 per million — 90% below fresh input, which is what a long system prompt resent on every call actually pays. It ranks #1459 of 1557 priced chat listings by input price, which makes it 8.3× the median priced model in this index ($0.6).

Three concrete workloads on this model's rates: a 20,000 requests/day support bot sending 2,500 tokens in and 400 out costs about $13500 a month; summarising 5,000 documents a day at 12,000 tokens in and a 600-token summary costs about $11250; and 1,000 agent sessions a day at 30,000 input and 5,000 output tokens costs about $8250. With 80% of that input served from cache it drops to $5010 — the cache rate matters more than the headline rate for this shape of workload.

How Amazon Bedrock bills: Amazon Bedrock bills per million tokens inside an AWS account, with regional endpoints and, for some models, batch inference priced differently from on-demand calls. The rate above is the one this index normalises to (USD per million tokens); the authoritative figure is always the provider's own pricing page, linked below.

The figure above excludes batch and committed-use discounts, regional rates, free tiers and enterprise agreements — providers publish those separately, and they can move the real bill by tens of percent. Context is listed as 1M tokens with at most 128K tokens of output per call; some providers bill requests above a long-context threshold at a higher tier, which this page does not model. Token counting also differs between providers' tokenizers, so two providers quoting the same rate can bill a different total for the same text.

Use Claude Opus 4 8 when you need vision input, tool calling, reasoning modes, prompt caching and the monthly figure above fits the budget. If it does not, Gemini Exp 1114 (Google Gemini) is the nearest cheaper option at free per million input tokens — keep the capability flags in mind before swapping, since a cheaper model that lacks one of the features above will cost more once you work around it.