Claude Sonnet 4.6

Provider: Anthropic US Official API price

Input price
$3.00
/ 1M tokens
Output price
$15.00
/ 1M tokens
Cache price
$0.30
/ 1M tokens (read)
Batch API
$1.50 / $7.50
input / output per 1M tokens
Context
1M tokens
Max output 128K tokens

Overview

Claude Sonnet 4.6 is a mid-tier model from Anthropic (US), designed for complex general-purpose tasks requiring extensive context, advanced capabilities, and a balance between performance and cost. It targets workloads that benefit from its sophisticated reasoning and multimodal understanding.

Strengths

  • Boasts an exceptionally large 1,000,000-token context window for processing vast amounts of information.
  • Offers multimodal capabilities, enabling the model to understand and process visual inputs.
  • Supports robust function-calling, allowing seamless integration with external tools and systems.
  • Provides a reliable JSON mode for structured and predictable output generation.
  • Features 'thinking' capability, indicating advanced reasoning and problem-solving skills.
  • Capable of generating very long and detailed responses, with a maximum output of 128,000 tokens.

Best use cases

  • Advanced document analysis and summarization across large codebases, legal documents, or research papers.
  • Developing intelligent agents that can interact with APIs and external tools through function-calling.
  • Automating complex workflows requiring structured data extraction from diverse inputs, including images.
  • Building applications that need highly structured, machine-readable output formats via JSON mode.
  • Complex reasoning tasks, strategic planning, and sophisticated content generation requiring deep understanding.

Pricing position

Claude Sonnet 4.6 maintains the same pricing as its predecessors, Sonnet 4 and Sonnet 4.5, at $3 per 1M input tokens and $15 per 1M output tokens. This positions it significantly above the more cost-optimized Claude Haiku series, which range from $0.25 to $1 per 1M input tokens. It is identically priced to several peer models like Grok 4 and Sonar Pro from other providers, placing it squarely in a competitive mid-to-high-tier segment.

Caveats

This model does not include a free tier, meaning developers will incur costs from the outset for experimentation and development. For simpler or less context-heavy tasks, Anthropic's more cost-effective Haiku models may be a better choice to optimize expenses. Its $3/M input and $15/M output token pricing makes it less suitable for high-volume, low-value content generation where cost efficiency per token is paramount.

Notes drafted by an LLM from the figures on this page (2026-05-21) and rewritten when the price changes.

Key facts

Price basis
Official API price
Max output
128K tokens
Listed on OpenRouter
2026-02-17
Tracked since
2026-05-07
Last updated
2026-11-10

Features

Image inputFunction callingJSON outputReasoning

Monthly cost estimate

Scale Input (M) Output (M) Monthly
0.1M (hobby) 0.1 0.05 $1.05
1M (small app) 1 0.5 $10.50
10M (mid scale) 10 5 $105.00
100M (large) 100 50 $1050.00

※ Cache and batch discounts not factored in. Detailed calculator

Official resources

Other models from Anthropic

Similar-priced models

Source: OpenRouter API (anthropic/claude-sonnet-4.6)