Claude Opus 4.5 vs GPT-5.2 Codex
Wins 3 of 8 areas
Agents · Facts · Math
Wins 5 of 8 areas
Coding · Reasoning · Images and charts · Long documents · Following instructions
GPT-5.2 Codex is the stronger all-rounder.Claude Opus 4.5 is better at facts.
Scores updated · 28 tests both models report · How we compare
Where each one wins
Tests won in each of the eight areas we test. Each piece is one test, so a longer bar means more evidence; grey means the two scored within a point of each other.
- ReasoningHard problems that need careful thinking13GPT-5.2 Codex3 of 4 tests
- Following instructionsDoing exactly what it is asked02GPT-5.2 Codex2 of 2 tests
- CodingWriting and fixing software34GPT-5.2 Codex4 of 7 tests
- Images and chartsUnderstanding pictures, charts and video01GPT-5.2 Codex1 of 1 test
- Long documentsFinding answers in very long texts01GPT-5.2 Codex1 of 1 test
- FactsGetting facts right instead of making them up20Claude Opus 4.52 of 2 tests
- AgentsCarrying out multi-step tasks on its own10Claude Opus 4.51 of 1 test
- MathCompetition and research-level math10Claude Opus 4.51 of 1 test
Agents, images and charts, math and long documents rest on a single test each.
What it costs
Prices per million tokens, roughly 750,000 words. The bars show the cost of a million tokens read plus a million written.
GPT-5.2 Codex costs 48% less for the same work.
The biggest differences
The three tests each model wins by the widest margin. Scores are out of 100.
Where GPT-5.2 Codex pulls ahead
- Follows unfamiliar, precisely checkable instructionsIFBench+19.6points ahead
- Fixes real GitHub issues, working as an agentLiveBench · Agentic Coding+9.7points ahead
- Very hard expert questions across many subjectsHumanity's Last Exam+5.6points ahead
Where Claude Opus 4.5 pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+12.4points ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+9.9points ahead
- Fixes real GitHub issues in Python projectsSWE-bench Verified+8.1points ahead
Every test, side by side
All 28 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingGPT-5.2 Codex
- LMArena · WebDevClaude Opus 4.5 by 129 rating points14681339+129 rating
- Terminal-Bench HardClaude Opus 4.5 by 9.94737.1+9.9
- LiveBench · Agentic CodingGPT-5.2 Codex by 9.739.749.4+9.7
- SWE-bench VerifiedClaude Opus 4.5 by 8.180.972.8+8.1
- SciCodeGPT-5.2 Codex by 5.149.554.6+5.1
- SWE-bench ProGPT-5.2 Codex by 4.45256.4+4.4
- LiveBench · CodingGPT-5.2 Codex by 3.979.783.6+3.9
ReasoningGPT-5.2 Codex
- Humanity's Last ExamGPT-5.2 Codex by 5.630.135.7+5.6
- CritPtGPT-5.2 Codex by 4.14.68.7+4.1
- GPQA DiamondGPT-5.2 Codex by 3.386.689.9+3.3
- LiveBench · ReasoningClaude Opus 4.5 by 2.480.177.7+2.4
FactsClaude Opus 4.5
- AA-Omniscience · Non-hallucinationClaude Opus 4.5 by 12.43926.6+12.4
- AA-Omniscience · AccuracyClaude Opus 4.5 by 5.546.641.1+5.5
Following instructionsGPT-5.2 Codex
- IFBenchGPT-5.2 Codex by 19.65877.6+19.6
- LiveBench · Instruction FollowingGPT-5.2 Codex by 3.962.566.5+3.9
Other results9 tests, not counted
Tests outside the eight areas. They are not counted above: several are summary scores built from other tests, or the same test under another name.
- AA-OmniscienceClaude Opus 4.5 by 16.214-2.2+16.2
- livebench_languageClaude Opus 4.5 by 7.681.373.7+7.6
- Artificial Analysis Coding IndexClaude Opus 4.5 by 4.847.843+4.8
- Terminal-Bench 2.0GPT-5.2 Codex by 4.759.364+4.7
- livebench_data_analysisGPT-5.2 Codex by 3.874.478.2+3.8
- AA Agentic IndexClaude Opus 4.5 by 3.159.656.5+3.1
- τ²-Bench Telecom (AA run)GPT-5.2 Codex by 2.689.592.1+2.6
- LiveBenchClaude Opus 4.5 by 1.77674.3+1.7
- AA Intelligencetie29.128.5tie
Questions people ask
Which is better, Claude Opus 4.5 or GPT-5.2 Codex?
GPT-5.2 Codex wins five of the eight areas we test: coding, reasoning, images and charts, long documents and following instructions. Claude Opus 4.5 wins agents, facts and math.
Which is better for coding?
GPT-5.2 Codex. It wins 4 of the 7 coding tests both models report; Claude Opus 4.5 wins 3.
Which is cheaper?
Claude Opus 4.5 costs $5.00 per million input tokens and $25.00 per million output tokens; GPT-5.2 Codex costs $1.75 and $14.00. That makes GPT-5.2 Codex about 48% cheaper for the same work.
How do you compare the two?
We use the 28 benchmark tests both models have published scores on. The verdict counts the 19 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 9 are listed but not counted, because several are summary scores or repeat a test. Each score is the one shown on the model's own page.