Router DeskHomeClaude Sonnet 5 pricing
Claude Sonnet 5 pricing
Sonnet 5 is cheaper than Sonnet 4.5 on OpenRouter and keeps a 1,000,000-token window.
Updated 2026-09-27
On 27 Sep 2026 openai-style pricing for anthropic/claude-sonnet-5 was $2 per million input tokens and $10 per million output tokens. Cache reads were $0.20 per million. Cache writes were $2.50, or $4 for the one-hour cache. Web search was $0.01. Batch was $1 input and $5 output. Context was 1,000,000 tokens.
The card calls Sonnet 5 Anthropic’s most capable Sonnet, with adaptive thinking and effort levels from low through max. It is the default to price against Haiku 4.5 ($1 / $5, 200K) and Opus 4.8 ($5 / $25, 1M).
A 100,000-token prompt and a 20,000-token answer is $0.20 + $0.20 = $0.40 on the standard rate. The same shape on Sonnet 4.5 is $0.30 + $0.30 = $0.60, before the long-context surcharge Sonnet 4.5 adds past 200K.
Questions
How much is Claude Sonnet 5?
$2 per million input tokens and $10 per million output tokens on 27 Sep 2026. Batch is $1 and $5.
What is the model id?
anthropic/claude-sonnet-5
Is Sonnet 5 cheaper than Sonnet 4.5?
Yes, on the rates checked that day: $2 / $10 versus $3 / $15, and Sonnet 4.5 charges more again above 200,000 prompt tokens.
Related
- Claude Haiku vs Sonnet
Haiku 4.5 is $1 / $5 per million tokens with a 200K context. Sonnet 4.5 is $3 / $15 and up to 1M. Sonnet 5 is $2 / $10.
- Opus 4 pricing
Claude Opus 4’s card is still on OpenRouter, with no endpoint on 27 Sep 2026. Live Opus rows that day ran from $5 / $25 to $15 / $75 per million tokens.
- Claude Code on OpenRouter
Point Claude Code at https://openrouter.ai/api, put the OpenRouter key in ANTHROPIC_AUTH_TOKEN, and set ANTHROPIC_API_KEY to an empty string.
- Coding models
Coding-model choice on OpenRouter is price, context, and tool-call reliability. DeepSeek V4, Qwen3.7 Max, GLM 5.2, and Sonnet 5 are the slugs to price first.