Claude Opus 4.8 vs GPT-4o
Wins 7 of 7 areas
Coding · Agents · Reasoning · Facts · Images and charts · Long documents · Following instructions
Wins 0 of 7 areas
—
Claude Opus 4.8 is the stronger all-rounder.GPT-4o is cheaper.
Scores updated · 21 tests both models report · How we compare
Where each one wins
Tests won in each of the seven areas where both have results. 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 thinking40Claude Opus 4.84 of 4 tests
- CodingWriting and fixing software30Claude Opus 4.83 of 3 tests
- AgentsCarrying out multi-step tasks on its own20Claude Opus 4.82 of 2 tests
- FactsGetting facts right instead of making them up21Claude Opus 4.82 of 3 tests
- Images and chartsUnderstanding pictures, charts and video10Claude Opus 4.81 of 1 test
- Long documentsFinding answers in very long texts10Claude Opus 4.81 of 1 test
- Following instructionsDoing exactly what it is asked10Claude Opus 4.81 of 1 test
Images and charts, long documents and following instructions 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-4o costs 33% less for the same work.
The biggest differences
The tests each model wins by the widest margin, up to three each. Scores are out of 100.
Where Claude Opus 4.8 pulls ahead
- Completes tasks by operating a computer desktopOSWorld-Verified+78.4points ahead
- Fixes real GitHub issues in Python projectsSWE-bench Verified+55.4points ahead
- Hard command-line tasks in a real terminalTerminal-Bench Hard+50points ahead
Where GPT-4o pulls ahead
- Avoids making up answers it doesn't knowAA-Omniscience · Non-hallucination+1.4points ahead
Every test, side by side
All 21 tests both models report. The winning score is in its model's colour; marks a score checked independently.
CodingClaude Opus 4.8
- SWE-bench VerifiedClaude Opus 4.8 by 55.488.633.2+55.4
- Terminal-Bench HardClaude Opus 4.8 by 5058.38.3+50
- SciCodeClaude Opus 4.8 by 21.154.433.3+21.1
AgentsClaude Opus 4.8
- OSWorld-VerifiedClaude Opus 4.8 by 78.483.45+78.4
- GDPValClaude Opus 4.8 by 47.847.80+47.8
ReasoningClaude Opus 4.8
- ARC-AGI-2Claude Opus 4.8 by 72.172.10+72.1
- Humanity's Last ExamClaude Opus 4.8 by 46.948.71.8+46.9
- GPQA DiamondClaude Opus 4.8 by 39.49252.6+39.4
- CritPtClaude Opus 4.8 by 20.920.90+20.9
FactsClaude Opus 4.8
- SimpleQA VerifiedClaude Opus 4.8 by 275326+27
- AA-Omniscience · AccuracyClaude Opus 4.8 by 25.148.823.7+25.1
- AA-Omniscience · Non-hallucinationGPT-4o by 1.460.762.1+1.4
Images and chartsClaude Opus 4.8
- CharXiv (RQ)Claude Opus 4.8 by 31.189.958.8+31.1
Following instructionsClaude Opus 4.8
- IFBenchClaude Opus 4.8 by 26.262.236+26.2
Other results6 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.
- ARC-AGI-1Claude Opus 4.8 by 8892.54.5+88
- τ²-Bench Telecom (AA run)Claude Opus 4.8 by 65.594.428.9+65.5
- Artificial Analysis Coding IndexClaude Opus 4.8 by 50.174.324.2+50.1
- AA-OmniscienceClaude Opus 4.8 by 39.328.8-10.5+39.3
- AA IntelligenceClaude Opus 4.8 by 34.541.87.3+34.5
- AA Agentic IndexClaude Opus 4.8 by 34.242.68.4+34.2
Questions people ask
Which is better, Claude Opus 4.8 or GPT-4o?
Claude Opus 4.8 wins all seven areas where both have results: coding, agents, reasoning, facts, images and charts, long documents and following instructions. GPT-4o wins none, but costs 33% less.
Which is better for coding?
Claude Opus 4.8. It wins 3 of the 3 coding tests both models report; GPT-4o wins none.
Which is cheaper?
Claude Opus 4.8 costs $5.00 per million input tokens and $25.00 per million output tokens; GPT-4o costs $5.00 and $15.00. That makes GPT-4o about 33% cheaper for the same work.
How do you compare the two?
We use the 21 benchmark tests both models have published scores on. The verdict counts the 15 tests in the eight capability areas, and a gap under one point (ten on rating-style scales) counts as a tie. The other 6 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.