Claude Sonnet 5 vs GPT-5

GPT-5 is cheaper for every workload modelled here.

List prices checked

Specifications and rates

PropertyClaude Sonnet 5GPT-5
Input per 1M$3.00$1.25
Output per 1M$15.00$10.00
Cached input per 1M$0.300$0.125
Blended (3:1)$6.00$3.44
Context window200K400K
Max output64K128K
Inputs acceptedtext, imagetext, image

Source: Anthropic pricing (August 2026) · OpenAI pricing (May 2026). Published list prices only — committed-use discounts, regional variation, and per-request charges are not included. Not yet independently re-checked against the vendor’s page.

Monthly cost by workload

Identical token volumes through both rate cards. Output-heavy profiles are where the ranking flips.

WorkloadClaude Sonnet 5GPT-5Cheaper
In-app chat assistantA support or help assistant handling a few thousand conversations a month, with short prompts and short answers.$135$75GPT-5saves $60/mo (44%)
RAG over documentsRetrieval-augmented answering where each request stuffs several retrieved passages into the prompt, so input dominates.$1,125$525GPT-5saves $600/mo (53%)
Coding agentAn agentic loop that reads files and writes patches, generating heavy output and re-sending large context on every turn.$2,700$1,500GPT-5saves $1,200/mo (44%)

What each one is for

Claude Sonnet 5

by Anthropic

The workhorse of the Claude family and the default for production work at volume. Introductory pricing of $2/$10 per million tokens runs to 31 August 2026; the $3/$15 standard rate shown here is what applies after that.

GPT-5

by OpenAI

OpenAI's flagship, priced aggressively below the other frontier models on input and with a 90% cached-input discount that makes long system prompts nearly free to reuse.

Common questions

Is Claude Sonnet 5 or GPT-5 cheaper in 2026?
GPT-5, for every workload modelled here. On a blended 3:1 basis it costs $3.44 per million tokens against Claude Sonnet 5's $6.00.
Which has the larger context window?
GPT-5, at 400K tokens against 200K. A larger window only helps if you can afford to fill it — the cost of one full-context request is in the table above.
Which is cheaper for output-heavy work like agents?
GPT-5, at $10.00 per million output tokens against $15.00. Reasoning tokens bill as output, so for agentic loops this axis usually decides the invoice regardless of what the input rates look like.
What is not included in these figures?
Batch-processing discounts, fine-tuning, long-context surcharges above a provider's standard tier, and any enterprise agreement. These are published pay-as-you-go list prices, checked against each provider's own pricing page on the date shown.
Is switching from Claude Sonnet 5 to GPT-5 worth it?
On price alone it is about 43% cheaper on a blended basis. Whether that survives contact with your workload depends on output quality and on retry rate — a cheaper model that needs two attempts is not cheaper. Model the specific volumes in the calculator before you migrate.

Other model comparisons