# Claude Opus 5 vs GPT-5.3 Codex

> Claude Opus 5 is the stronger model overall, scoring 67.7 to 45.5 on the Noometry Index. GPT-5.3 Codex costs 2.1× less per token, which makes it the better buy when Claude Opus 5's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/claude-opus-5-vs-gpt-5-3-codex
- Last updated: 2026-10-10
- Shared benchmarks: 5

## Summary

- They share 5 benchmarks with published results for both. Claude Opus 5 scores higher in 2 categories and GPT-5.3 Codex in 0 categories; 2 gaps are clear of the uncertainty.
- The widest gap is in coding, where Claude Opus 5 leads 67.0 to 48.5.
- The biggest single-benchmark swing is WeirdML: 91.8% for Claude Opus 5 and 79.3% for GPT-5.3 Codex.
- GPT-5.3 Codex is cheaper at $1.75 / $14 per million input/output tokens, against $5 / $25 for Claude Opus 5.
- Claude Opus 5 accepts more context: 1M tokens versus 400K.

## Snapshot

| | Claude Opus 5 | GPT-5.3 Codex |
|---|---|---|
| Provider | Anthropic | OpenAI |
| Noometry Index | 67.7 | 45.5 |
| Rank | 4 | 61 |
| Context | 1M | 400K |
| Input $/M | $5 | $1.75 |
| Output $/M | $25 | $14 |
| Weights | Proprietary | Proprietary |

## Coding

- Claude Opus 5: 67.0 (#5)
- GPT-5.3 Codex: 48.5 (#51)

| Benchmark | Claude Opus 5 | GPT-5.3 Codex |
|---|---|---|
| LMArena WebDev | 1694 | 1408 |
| WeirdML | 91.8% | 79.3% |
| ALE-Bench | 2,165 | 1,655 |
| SWE-bench Verified | — | 74.8% |
| DeepSWE | 73.6% | — |
| FrontierCode | 53.4% | — |
| CursorBench | 46.6% | — |
| FrontierSWE | 52% | — |
| SciCode | 56.4% | — |
| LMArena Coding | 1533 | — |

## Agentic & Tool Use

- Claude Opus 5: 55.5 (#1)
- GPT-5.3 Codex: 47.8 (#9)

| Benchmark | Claude Opus 5 | GPT-5.3 Codex |
|---|---|---|
| Vending-Bench 2 | 11,182 | 5,940 |
| Terminal-Bench | — | 78.4% |
| APEX-Agents | 65.8% | — |
| OSWorld 2.0 | 31.4% | — |
| τ²-bench Banking | 48.7% | — |
| PostTrainBench | 35% | — |
| BALROG | 63.4% | — |
| GBAEval | 79.6% | — |
| GDP.pdf | 24% | — |
| METR Time Horizons | — | 74.5% |

## Reasoning

- Claude Opus 5: 77.1 (#3)
- GPT-5.3 Codex: —

| Benchmark | Claude Opus 5 | GPT-5.3 Codex |
|---|---|---|
| Epoch Capabilities Index | 162.67 | 156.84 |
| ARC-AGI-2 | 90.4% | — |
| SimpleBench | 80.6% | — |
| NYT Connections (extended) | 94.3% | — |
| ARC-AGI-1 | 97.5% | — |
| CritPt | 29.1% | — |
| Chess Puzzles | 42% | — |
| EBR-Bench | 45.7% | — |
| LMArena Hard Prompts | 1525 | — |
| Mystery Game Puzzles | 59% | — |
| DTBench | 97.9% | — |
| LMCA | 64.5% | — |
| Bench to the Future 3 | 0.12 | — |

## Math

- Claude Opus 5: 86.1 (#6)
- GPT-5.3 Codex: —

| Benchmark | Claude Opus 5 | GPT-5.3 Codex |
|---|---|---|
| FrontierMath (Tiers 1-3) | 85.6% | — |
| FrontierMath Tier 4 | 73.2% | — |
| OTIS Mock AIME 2024-2025 | 98.9% | — |
| ProofBench | 99% | — |
| LMArena Math | 1533 | — |

## Knowledge

- Claude Opus 5: 66.8 (#8)
- GPT-5.3 Codex: —

| Benchmark | Claude Opus 5 | GPT-5.3 Codex |
|---|---|---|
| GPQA Diamond | 93.9% | — |
| SimpleQA Verified | 59.9% | — |
| LMArena Expert | 1558 | — |

## Multimodal

- Claude Opus 5: 51.1 (#6)
- GPT-5.3 Codex: —

| Benchmark | Claude Opus 5 | GPT-5.3 Codex |
|---|---|---|
| LMArena Vision | 1321 | — |
| Blueprint-Bench 2 | 30.4% | — |
| Furniture Assembly | 60.8% | — |
| LMArena Document | 1516 | — |

## Multilingual

- Claude Opus 5: 58.8 (#3)
- GPT-5.3 Codex: —

| Benchmark | Claude Opus 5 | GPT-5.3 Codex |
|---|---|---|
| LMArena Non-English | 1500 | — |
| LMArena Chinese | 1575 | — |
| LMArena French | 1522 | — |
| LMArena German | 1522 | — |
| LMArena Japanese | 1510 | — |
| LMArena Korean | 1526 | — |
| LMArena Russian | 1507 | — |
| LMArena Spanish | 1519 | — |

## Instruction Following

- Claude Opus 5: 79.3 (#5)
- GPT-5.3 Codex: —

| Benchmark | Claude Opus 5 | GPT-5.3 Codex |
|---|---|---|
| LMArena Instruction Following | 1516 | — |

## Long Context

- Claude Opus 5: 46.5 (#20)
- GPT-5.3 Codex: —

| Benchmark | Claude Opus 5 | GPT-5.3 Codex |
|---|---|---|
| LMArena Longer Query | 1516 | — |

## Writing & Preference

- Claude Opus 5: 79.3 (#2)
- GPT-5.3 Codex: —

| Benchmark | Claude Opus 5 | GPT-5.3 Codex |
|---|---|---|
| LMArena Text | 1506 | — |
| LMArena Creative Writing | 1491 | — |
| EQ-Bench Creative Writing | 2133 | — |
| EQ-Bench 4 | 1385 | — |
| LMArena Multi-Turn | 1496 | — |

## FAQ

### Is Claude Opus 5 better than GPT-5.3 Codex?

Claude Opus 5 is the stronger model overall, scoring 67.7 to 45.5 on the Noometry Index. GPT-5.3 Codex costs 2.1× less per token, which makes it the better buy when Claude Opus 5's lead doesn't matter for your workload.

### Which is cheaper, Claude Opus 5 or GPT-5.3 Codex?

GPT-5.3 Codex is cheaper. It lists at $1.75 per million input tokens and $14 per million output tokens; Claude Opus 5 lists at $5 and $25.

### Is Claude Opus 5 or GPT-5.3 Codex better for coding?

Claude Opus 5 scores higher on coding benchmarks: 67.0 versus 48.5 in the Noometry coding category.

### Which has the bigger context window?

Claude Opus 5 does, with 1M tokens against 400K.

### How many benchmarks do Claude Opus 5 and GPT-5.3 Codex share?

5 benchmarks have published results for both models. Claude Opus 5 has 57 scored results on Noometry and GPT-5.3 Codex has 8.