Compare

Claude Haiku 4.5 vs o4

Anthropic vs OpenAI. Specs, benchmarks, and real per-task cost — all in one page.

Verdict

Claude Haiku 4.5 is ~16.4× cheaper on a 3:1 input:output blend.

Anthropic

Fast, cheap, and smart enough for most routing and extraction tasks. Great base model for subagents.

Sub-second latency10x cheaper than OpusStrong structured output

OpenAI

Deep-reasoning model. Spends more tokens thinking to crush math, science, and hard coding problems.

State-of-the-art math and scienceExtremely high-ceiling reasoning

Pricing

Input / 1M$0.80$15.00

Output / 1M$4.00$60.00

Context200K200K

Max output16K100K

LicenseProprietaryProprietary

Released2025-10-012026-02-18

Benchmarks

LMSYS ELO1305.0—

MMLU Pro81.493.8

HumanEval87.092.3

SWE-Bench49.268.0

MATH76.896.2

GPQA—82.5

IFEval87.3—

Per-task cost

Summarize a 1-hour meeting transcript$0.014$0.255

Review a 500-line pull request$0.014$0.240

Answer a customer support ticket$0.0056$0.093

Extract structured data from a resume$0.0080$0.135

Debug a stack trace with context$0.013$0.210

Per-call cost using published token counts for each task. Real-world prompts vary.

Add up to four models, tweak tasks live.