Skip to model information

Claude Opus 4.6

Legacy · available

A legacy, still-available Opus release with adaptive thinking and text-and-image input.

Access details: Anthropic · Claude API

The model ID to use in API requests.

claude-opus-4-6
Updated Sep 23, 2026

Capabilities & limits

Context window
1M
Maximum output
128K
Image input
Supported
Tool calling
Supported
Structured output
Supported

Pricing

USD / 1M tokens
Input
$5.00
Output
$25.00

USD per million tokens. Standard global text requests, excluding cache, batch, tools, taxes, and regional premiums.

Full rate card 5 rates

All prices in USD.

Base input tokens

$5.00 per 1M tokens
  • Claude API global standard

Output tokens

$25.00 per 1M tokens
  • Claude API global standard

Cache hits and refreshes

$0.50 per 1M tokens
  • prompt caching

5 minute cache writes

$6.25 per 1M tokens
  • prompt caching
  • 5 minute duration

1 hour cache writes

$10.00 per 1M tokens
  • prompt caching
  • 1 hour duration

Full 1M-token context window uses standard pricing, including requests over 200K input tokens; prompt caching and Batch discounts remain at standard rates.

Price history
Sep 13, 2026

Input $5.00 per million tokens

Oct 20, 2026

Input $5.00 per million tokens

Sep 13, 2026

Output $25.00 per million tokens

Oct 20, 2026

Output $25.00 per million tokens

Technical details

Parameters
Not published
Active parameters
Not published
Architecture
Not published
Training / knowledge cutoff
Aug 2025
Input modalities
Text, image
Output modalities
Text

Release & lifecycle

First available
Feb 5, 2026
legacy
As of Sep 13, 2026
legacy
As of Oct 20, 2026

Considering a switch?

Specs only go so far. Try both on your own tasks and compare quality, speed and cost.

Compare with GPT-Rosalind →

Prices and availability as of October 20, 2026. Anthropic's own pages have the latest.