# Claude Opus 4.5 vs Grok 4.3

> Claude Opus 4.5 is the stronger model overall, scoring 50.5 to 43.6 on the Noometry Index. Grok 4.3 costs 6.4× less per token, which makes it the better buy when Claude Opus 4.5's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/claude-opus-4-5-vs-grok-4-3
- Last updated: 2026-10-10
- Shared benchmarks: 34

## Summary

- They share 34 benchmarks with published results for both. Claude Opus 4.5 scores higher in 8 categories and Grok 4.3 in 2 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in agentic & tool use, where Claude Opus 4.5 leads 47.2 to 27.6.
- The biggest single-benchmark swing is ProofBench: 36% for Claude Opus 4.5 and 11% for Grok 4.3.
- Grok 4.3 is cheaper at $1.25 / $2.50 per million input/output tokens, against $5 / $25 for Claude Opus 4.5.
- Grok 4.3 accepts more context: 1M tokens versus 200K.

## Snapshot

| | Claude Opus 4.5 | Grok 4.3 |
|---|---|---|
| Provider | Anthropic | xAI |
| Noometry Index | 50.5 | 43.6 |
| Rank | 39 | 80 |
| Context | 200K | 1M |
| Input $/M | $5 | $1.25 |
| Output $/M | $25 | $2.50 |
| Weights | Proprietary | Proprietary |

## Coding

- Claude Opus 4.5: 55.4 (#21)
- Grok 4.3: 41.3 (#117)

| Benchmark | Claude Opus 4.5 | Grok 4.3 |
|---|---|---|
| LMArena WebDev | 1493 | 1357 |
| WeirdML | 63.7% | 49.9% |
| LMArena Coding | 1504 | 1415 |
| ALE-Bench | 1,025 | 944.17 |
| SWE-bench Verified | 76.7% | — |
| SWE-bench Verified (bash only) | 76.8% | — |
| SWE-bench Multilingual | 70.7% | — |
| SciCode | — | 47.3% |
| GSO | 26.5% | — |
| AlgoTune | 1.77 | — |

## Agentic & Tool Use

- Claude Opus 4.5: 47.2 (#11)
- Grok 4.3: 27.6 (#93)

| Benchmark | Claude Opus 4.5 | Grok 4.3 |
|---|---|---|
| LMArena Search | 1180 | 1165 |
| Vending-Bench 2 | 4,967 | 35.26 |
| Terminal-Bench | 63.1% | — |
| Berkeley Function Calling Leaderboard | 77.5% | — |
| GDPval | 45.5% | — |
| Remote Labor Index | 3.8% | — |
| τ²-bench Airline | 84% | — |
| τ²-bench Banking | 24.7% | — |
| τ²-bench Retail | 79.6% | — |
| τ²-bench Telecom | 92.3% | — |
| Cybench | 82% | — |
| DeepResearch Bench | 54.8% | — |
| OSWorld | 66.3% | — |
| BALROG | 43.5% | — |
| GDP.pdf | — | 8% |
| METR Time Horizons | 75% | — |

## Reasoning

- Claude Opus 4.5: 42.5 (#43)
- Grok 4.3: 34.9 (#60)

| Benchmark | Claude Opus 4.5 | Grok 4.3 |
|---|---|---|
| NYT Connections (extended) | 52.5% | 55.2% |
| Chess Puzzles | 12% | 25% |
| LMArena Hard Prompts | 1476 | 1397 |
| DTBench | 89.9% | 90.7% |
| LMCA | 44.5% | 38.3% |
| Epoch Capabilities Index | 150.1 | 149.13 |
| ForecastBench | 60.7 | 60.3 |
| ARC-AGI-2 | 37.6% | — |
| SimpleBench | 62% | — |
| Kagi LLM Benchmark | 80.2% | — |
| ARC-AGI-1 | 80% | — |
| CritPt | — | 8% |
| EnigmaEval | 11.9% | — |
| EBR-Bench | 14.3% | — |
| Mystery Game Puzzles | 22% | — |

## Math

- Claude Opus 4.5: 38.5 (#85)
- Grok 4.3: 45.8 (#60)

| Benchmark | Claude Opus 4.5 | Grok 4.3 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 34.4% | 42.8% |
| FrontierMath Tier 4 | 4.9% | 14.6% |
| OTIS Mock AIME 2024-2025 | 86.1% | 93.3% |
| ProofBench | 36% | 11% |
| LMArena Math | 1463 | 1388 |
| FrontierMath (Feb 2025 set) | 20.7% | — |
| FrontierMath Tier 4 (v1) | 4.2% | — |

## Knowledge

- Claude Opus 4.5: 56.4 (#38)
- Grok 4.3: 52.5 (#57)

| Benchmark | Claude Opus 4.5 | Grok 4.3 |
|---|---|---|
| GPQA Diamond | 86% | 88.8% |
| SimpleQA Verified | 45.7% | 33.2% |
| LMArena Expert | 1488 | 1385 |
| Humanity's Last Exam | 25.2% | — |
| Vectara Hallucination Rate | 10.9% | — |

## Multimodal

- Claude Opus 4.5: 31.1 (#99)
- Grok 4.3: 31.6 (#96)

| Benchmark | Claude Opus 4.5 | Grok 4.3 |
|---|---|---|
| LMArena Vision | — | 1231 |
| GeoBench | 75% | — |
| VPCT | 40% | — |
| Blueprint-Bench 2 | — | 0% |
| Furniture Assembly | 28.3% | — |
| LMArena Document | 1462 | — |

## Multilingual

- Claude Opus 4.5: 54.5 (#45)
- Grok 4.3: 50.7 (#114)

| Benchmark | Claude Opus 4.5 | Grok 4.3 |
|---|---|---|
| LMArena Non-English | 1438 | 1385 |
| LMArena Chinese | 1470 | 1423 |
| LMArena French | 1471 | 1412 |
| LMArena German | 1449 | 1395 |
| LMArena Japanese | 1415 | 1379 |
| LMArena Korean | 1424 | 1357 |
| LMArena Russian | 1448 | 1399 |
| LMArena Spanish | 1455 | 1402 |

## Instruction Following

- Claude Opus 4.5: 77.7 (#20)
- Grok 4.3: 72.3 (#135)

| Benchmark | Claude Opus 4.5 | Grok 4.3 |
|---|---|---|
| LMArena Instruction Following | 1478 | 1366 |

## Long Context

- Claude Opus 4.5: 46.4 (#21)
- Grok 4.3: 42.5 (#118)

| Benchmark | Claude Opus 4.5 | Grok 4.3 |
|---|---|---|
| LMArena Longer Query | 1480 | 1393 |
| CL-bench | 21.1% | — |

## Writing & Preference

- Claude Opus 4.5: 68.2 (#28)
- Grok 4.3: 58.6 (#112)

| Benchmark | Claude Opus 4.5 | Grok 4.3 |
|---|---|---|
| LMArena Text | 1451 | 1397 |
| LMArena Creative Writing | 1445 | 1380 |
| LMArena Multi-Turn | 1466 | 1406 |
| EQ-Bench Creative Writing | 1687 | — |
| EQ-Bench 4 | — | 1075 |

## FAQ

### Is Claude Opus 4.5 better than Grok 4.3?

Claude Opus 4.5 is the stronger model overall, scoring 50.5 to 43.6 on the Noometry Index. Grok 4.3 costs 6.4× less per token, which makes it the better buy when Claude Opus 4.5's lead doesn't matter for your workload.

### Which is cheaper, Claude Opus 4.5 or Grok 4.3?

Grok 4.3 is cheaper. It lists at $1.25 per million input tokens and $2.50 per million output tokens; Claude Opus 4.5 lists at $5 and $25.

### Is Claude Opus 4.5 or Grok 4.3 better for coding?

Claude Opus 4.5 scores higher on coding benchmarks: 55.4 versus 41.3 in the Noometry coding category.

### Which has the bigger context window?

Grok 4.3 does, with 1M tokens against 200K.

### How many benchmarks do Claude Opus 4.5 and Grok 4.3 share?

34 benchmarks have published results for both models. Claude Opus 4.5 has 69 scored results on Noometry and Grok 4.3 has 40.