Claude Opus 4.8 API Pricing Calculator

Compare what your prompts cost across every major LLM — with exact token counts for OpenAI models, right in your browser.

Prices verified 2026-07-24 · per 1M tokens · sources linked per provider

Input / 1M tokens

$5

Cached input / 1M

$0.5

Output / 1M tokens

$25

Context window: 1,000,000 tokensAPI model ID: claude-opus-4-8

At Claude Opus 4.8's rates ($5 per 1M input tokens, $25 per 1M output), a call with 1,000 input and 500 output tokens costs $0.0175. At 10,000 requests a month, that workload runs $175.00. A heavier call — a 100,000-token prompt returning 2,000 tokens — costs $0.55.

Within Anthropic Claude's lineup, Claude Opus 4.8 runs 5.0× the per-call cost of Claude Haiku 4.5 and 2.0× less than Claude Fable 5 at this workload. It is priced identically to Claude Opus 5 ($5 in / $25 out per 1M tokens) — for this calculator the two differ only in the model itself, not the bill.

Cache-read input is billed at $0.5 per 1M tokens — 90% below the standard input rate, so a long system prompt reused across calls mostly bills at the discounted rate.

Claude Opus 4.8's context window is 1,000,000 tokens.

ModelCost / callCost / month× best price
Claude Haiku 4.5cheapest$0.006$6.00
Claude Sonnet 5$0.012$12.002.0×
Claude Opus 5$0.03$30.005.0×
Claude Opus 4.8$0.03$30.005.0×
Claude Fable 5$0.06$60.0010.0×

More Anthropic Claude models: Claude Opus 5 · Claude Sonnet 5 · Claude Haiku 4.5 · Claude Fable 5 — or all Anthropic Claude pricing

Token counter

0 tokens

Token counts are exact for OpenAI models (tiktoken o200k_base, run in your browser). For other providers the count is an estimate — notably Claude Opus 4.7+/Sonnet 5 and similar newer models use a tokenizer that produces roughly 30% more tokens for the same text.

Related tools

Scaling AI content or ops? Elegant Atomics builds growth systems for B2B SaaS — Trevor’s company. Work with us →