Model ComparisonAnthropic

Claude Sonnet 5

Claude Sonnet 5 — the best balance of capability and cost in Anthropic's lineup.

VisionTool usePrompt cachingReasoningStandard
Pricing (verified 2026-08-19)
TypePrice / 1M tokensNotes
Input$3.00Per 1M input tokens
Output$15.00Per 1M output tokens
Cache read$0.30Cached prefix read-back
Cache write$3.75Anthropic 5m TTL; 2× for 1h TTL
Batch API$1.50 in / $7.50 out~50% off, async (up to 24h)
Source: Anthropic official pricing page · Last verified 2026-08-19
Context Window
200K tokens
150K words · 1 pages of text

At $3/M input and $15/M output, Sonnet 5 delivers near-Opus quality for most professional tasks.

Best for

  • Production chatbots and customer support
  • Code generation and debugging
  • Summarization at scale
  • RAG pipeline generation

Not ideal for

  • Absolute maximum reasoning quality
  • Cost-sensitive classification at millions of requests

Real cost scenarios

ScenarioEst. monthly costBreakdown
1M requests/month — 1k in, 300 out$7,500$3,000 input + $4,500 output
Customer support chatbot — 50k conv/month$4,50010 turns avg, 500 tokens/turn

Estimates with typical token distributions. Use the Prompt Cost Calculator for your exact workload.

How to cut costs on Claude Sonnet 5

  1. Cache your system prompt — at $3/M input and 85% hit rate, a 3,000-token system prompt saves ~$7.50 per 1,000 calls.
  2. For simple queries in your workload, route to Haiku 4.5 — 3-4× cheaper with similar quality on classification tasks.

Calculate your Claude Sonnet 5 costs

Prompt Cost Calculator
Exact cost for your prompt
Agent Cost Calculator
Multi-step agent loops
Prompt Caching Calculator
Caching savings
Conversation Cost
Chat history strategies

Other Anthropic models

Claude Opus 5$5.00/MClaude Haiku 4.5$1.00/MClaude Haiku 3.5$0.80/M

Compare Claude Sonnet 5

claude sonnet 5 vs gpt 5 4claude sonnet 5 vs claude opus 5gpt 5 mini vs claude haiku 4 5

Frequently Asked Questions

How does Claude Sonnet 5 compare to GPT-5.4?

On most coding and reasoning benchmarks they are within 2-3%. Sonnet 5 has a larger context window (200K vs 128K) and better document analysis. GPT-5.4 has a larger ecosystem of third-party integrations. For most production workloads, choose based on which API fits your infrastructure better.

Does Claude Sonnet 5 support tool/function calling?

Yes — full tool use support including parallel tool calls, forced tool choice, and streaming tool call results. Compatible with most agent frameworks including LangChain, LangGraph, and direct SDK use.

Pricing verified 2026-08-19 from Anthropic official pricing. All prices in USD per 1M tokens. See methodology and price changelog.