Provider

Anthropic API Pricing Claude Opus 4, Sonnet 4, Haiku

Anthropic builds the Claude family of AI assistants, known for their strong instruction-following, safety, and extended thinking capabilities. Their lineup spans from the lightweight Claude 3 Haiku to the frontier-class Claude Opus 4.

All Anthropic Models

10 models · Prices in USD per 1M tokens

Last verified: 2026-02-20
ModelTierInput $/1MOutput $/1M
Claude Opus 4.6Flagship$5.00$25.00
Claude Sonnet 4.6Mid-tier$3.00$15.00
Claude Opus 4.5Flagship$5.00$25.00
Claude Sonnet 4.5Mid-tier$3.00$15.00
Claude Haiku 4.5Budget$1.00$5.00
Claude Opus 4Flagship$15.00$75.00
Claude Sonnet 4Mid-tier$3.00$15.00
Claude 3.5 Sonnet v2Mid-tier$3.00$15.00
Claude 3.5 HaikuBudget$0.80$4.00
Claude 3 HaikuBudget$0.25$1.25

About Anthropic's Claude Models

Anthropic is an AI safety company founded in 2021 by former OpenAI researchers. Their Claude models are designed with a focus on being helpful, harmless, and honest, incorporating Constitutional AI training techniques that set them apart in the industry.

Claude Opus 4 is Anthropic's most capable model, excelling at complex analysis, extended reasoning, and nuanced writing tasks. At $15/$75 per million input/output tokens, it sits at the premium end but delivers top-tier benchmark performance. Claude Sonnet 4 offers an excellent balance of intelligence and cost at $3/$15, with a 64K max output window — the largest in the lineup.

For cost-sensitive workloads, the Haiku models offer strong performance at a fraction of the cost. Claude 3 Haiku remains one of the most affordable options at $0.25/$1.25 per million tokens. All Anthropic models support prompt caching, which can dramatically reduce input costs for repetitive system prompts and few-shot examples.

Disclosure: Pricing data on this page is sourced from Anthropic's official pricing page and was last verified on 2026-02-20. Prices are subject to change without notice. Some links on this site may be affiliate links. We may earn a commission at no additional cost to you. This does not influence our data or recommendations — all pricing is sourced directly from provider documentation.