TokenTally

Anthropic Claude

Opus 4.6 token costs

Use this guide to model the $5/M input, $25/M output rates and the 1M-token context window (beta).

Last pricing check: Mar 13, 2026

$5.00 per 1M prompt tokens$25.00 per 1M completion tokens1,000,000 token context

Why teams choose this model

Autonomous engineering copilots
Long-running agent teams
Complex planning + analysis
Open main calculator

Scenario planning

Realistic cost examples

Numbers use Claude Opus 4.6 pricing

Multi-agent dev pod

Claude Code orchestrates agent teams refactoring services with context compaction.

Per request

$0.0655

Per month

$314.40

Tokens sent

22,560,000

2600 prompt tokens2100 completion tokens4,800 requests/mo

Strategic analyst

Ops teams push long-form dossiers into Opus for extended thinking recommendations.

Per request

$0.051

Per month

$331.50

Tokens sent

24,700,000

2200 prompt tokens1600 completion tokens6,500 requests/mo

Agentic ETL

Opus agents fetch, filter, and summarize live data with context compaction enabled.

Per request

$0.0445

Per month

$320.40

Tokens sent

23,760,000

1900 prompt tokens1400 completion tokens7,200 requests/mo

Compare with

Anthropic

Claude Sonnet 4.6 Cost Planner

Updated Mar 13, 2026

Anthropic

Claude 3.7 Haiku Cost Calculator

Updated Mar 13, 2026

Anthropic

Claude 3.7 Sonnet Pricing Guide

Updated Mar 13, 2026

FAQs

How do I access the 1M-token context?

It’s currently in beta via the Claude API/Developer Platform. Long-context pricing applies once you exceed 200K tokens, per Anthropic’s docs.

What about prompt caching?

Prompt cache reads drop to $0.50/M tokens. Writes cost $6.25/M—the pricing page spells out both numbers for Opus 4.6.

Does Opus sit on the priority tier?

Yes—Anthropic lists Opus 4.6 in the Priority tier, so you can reserve capacity for high-SLA workloads.

Pricing sources