Comparison

Claude Haiku 4.5 vs GPT-4o

Per input token, Claude Haiku 4.5 runs 60% cheaper than GPT-4o. On generated (output) tokens, GPT-4o costs 50% more than Claude Haiku 4.5. Claude Haiku 4.5 has the larger context window than GPT-4o.

FieldClaude Haiku 4.5
GPT-4o
Legacy
ProviderAnthropicOpenAI
Context window200K128K
Max output64K
Input / 1M$1.0060% cheaper$2.50
Output / 1M$5.0050% cheaper$10.00
Cached input / 1M$0.1092% cheaper$1.25
Batch discount50%
Release date2025-10-01

Worked cost examples

Estimated monthly cost on three common usage shapes

WorkloadClaude Haiku 4.5GPT-4o

Chatbot

800 input / 200 output tokens per request, 100,000 requests/month

$180.00cheaper$400.00

RAG pipeline

6,000 input / 700 output tokens per request, 30,000 requests/month

$285.00cheaper$660.00

Batch summarization

20,000 input / 1,500 output tokens per request, 5,000 requests/month

$137.50cheaper$325.00

Claude Haiku 4.5 vs GPT-4o FAQ

Is Claude Haiku 4.5 cheaper than GPT-4o?
Claude Haiku 4.5 costs $1.00 per 1M input tokens versus $2.50 for GPT-4o — Claude Haiku 4.5 is 60% cheaper per input token.
Which is cheaper for output tokens, Claude Haiku 4.5 or GPT-4o?
Claude Haiku 4.5 charges $5.00 per 1M output tokens; GPT-4o charges $10.00 per 1M output tokens.
Which has the larger context window, Claude Haiku 4.5 or GPT-4o?
Claude Haiku 4.5 does, with a 200,000-token context window.
Do Claude Haiku 4.5 and GPT-4o offer cache pricing?
Claude Haiku 4.5 discounts cached input to $0.10 per 1M tokens. GPT-4o discounts cached input to $1.25 per 1M tokens.

Looking for another matchup? See all comparisons.