# GPT-5.5 vs Grok 4.20 (Non-Reasoning)

> GPT-5.5 is the stronger model overall, scoring 63.4 to 48.5 on the Noometry Index. Grok 4.20 (Non-Reasoning) costs 7.2× less per token, which makes it the better buy when GPT-5.5's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/gpt-5-5-vs-grok-4-20
- Last updated: 2026-10-10
- Shared benchmarks: 44

## Summary

- They share 44 benchmarks with published results for both. GPT-5.5 scores higher in 10 categories and Grok 4.20 (Non-Reasoning) in 0 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-5.5 leads 81.5 to 48.0.
- The biggest single-benchmark swing is FrontierMath Tier 4: 72.5% for GPT-5.5 and 17.1% for Grok 4.20 (Non-Reasoning).
- Grok 4.20 (Non-Reasoning) is cheaper at $1.25 / $2.50 per million input/output tokens, against $5 / $30 for GPT-5.5.
- GPT-5.5 accepts more context: 1.05M tokens versus 1M.

## Snapshot

| | GPT-5.5 | Grok 4.20 (Non-Reasoning) |
|---|---|---|
| Provider | OpenAI | xAI |
| Noometry Index | 63.4 | 48.5 |
| Rank | 8 | 48 |
| Context | 1.05M | 1M |
| Input $/M | $5 | $1.25 |
| Output $/M | $30 | $2.50 |
| Weights | Proprietary | Proprietary |

## Coding

- GPT-5.5: 58.3 (#13)
- Grok 4.20 (Non-Reasoning): 41.8 (#104)

| Benchmark | GPT-5.5 | Grok 4.20 (Non-Reasoning) |
|---|---|---|
| LMArena WebDev | 1512 | 1374 |
| WeirdML | 84.9% | 52.3% |
| LMArena Coding | 1494 | 1458 |
| ALE-Bench | 1,943 | 1,150 |
| SWE-bench Verified | 80.6% | — |
| DeepSWE | 67% | — |
| FrontierCode | 43% | — |
| SciCode | 56.1% | — |
| GSO | 40.2% | — |
| MirrorCode | 10% | — |

## Agentic & Tool Use

- GPT-5.5: 50.7 (#6)
- Grok 4.20 (Non-Reasoning): 34.3 (#43)

| Benchmark | GPT-5.5 | Grok 4.20 (Non-Reasoning) |
|---|---|---|
| Terminal-Bench | 84.7% | 57.3% |
| τ²-bench Banking | 44.6% | 18% |
| LMArena Search | 1242 | 1189 |
| Vending-Bench 2 | 7,524 | 4,663 |
| APEX-Agents | 55.1% | — |
| OSWorld 2.0 | 13% | — |
| Remote Labor Index | 6.3% | — |
| DeepResearch Bench | 54% | — |
| PostTrainBench | 27.2% | — |
| ExploitBench | 47.4% | — |
| GBAEval | 53.2% | — |
| GDP.pdf | 26% | — |

## Reasoning

- GPT-5.5: 72.7 (#8)
- Grok 4.20 (Non-Reasoning): 52.2 (#30)

| Benchmark | GPT-5.5 | Grok 4.20 (Non-Reasoning) |
|---|---|---|
| ARC-AGI-2 | 85% | 65.1% |
| Kagi LLM Benchmark | 88.8% | 75% |
| NYT Connections (extended) | 96.2% | 85.4% |
| ARC-AGI-1 | 95% | 89.5% |
| Chess Puzzles | 54% | 24% |
| LMArena Hard Prompts | 1489 | 1451 |
| DTBench | 96% | 90.1% |
| LMCA | 54.3% | 38.7% |
| Epoch Capabilities Index | 159.25 | 152.01 |
| ForecastBench | 60.6 | 61.4 |
| SimpleBench | 69% | — |
| CritPt | 27.1% | — |
| Thematic Generalization | — | 63.8% |
| EBR-Bench | 34.3% | — |
| Mystery Game Puzzles | 56% | — |
| Surface Evolver Bench | 88.1% | — |
| Bench to the Future 3 | 0.14 | — |

## Math

- GPT-5.5: 81.5 (#9)
- Grok 4.20 (Non-Reasoning): 48.0 (#54)

| Benchmark | GPT-5.5 | Grok 4.20 (Non-Reasoning) |
|---|---|---|
| FrontierMath (Tiers 1-3) | 85.3% | 44.9% |
| FrontierMath Tier 4 | 72.5% | 17.1% |
| OTIS Mock AIME 2024-2025 | 100% | 92.2% |
| ProofBench | 50% | 14% |
| LMArena Math | 1486 | 1455 |
| MathArena Final-Answer Competitions | 94.3% | — |
| FrontierMath (Feb 2025 set) | 51.7% | — |
| FrontierMath Erdős | 0% | — |
| FrontierMath Tier 4 (v1) | 35.4% | — |

## Knowledge

- GPT-5.5: 64.4 (#14)
- Grok 4.20 (Non-Reasoning): 52.8 (#54)

| Benchmark | GPT-5.5 | Grok 4.20 (Non-Reasoning) |
|---|---|---|
| GPQA Diamond | 94% | 89.3% |
| SimpleQA Verified | 63% | 30.2% |
| LMArena Expert | 1508 | 1440 |
| Vectara Hallucination Rate | 9.3% | — |

## Multimodal

- GPT-5.5: 47.1 (#11)
- Grok 4.20 (Non-Reasoning): 33.2 (#88)

| Benchmark | GPT-5.5 | Grok 4.20 (Non-Reasoning) |
|---|---|---|
| LMArena Vision | 1297 | 1263 |
| Blueprint-Bench 2 | 36.2% | 0% |
| LMArena Document | 1486 | 1416 |
| Furniture Assembly | 44.2% | — |

## Multilingual

- GPT-5.5: 56.5 (#19)
- Grok 4.20 (Non-Reasoning): 54.7 (#38)

| Benchmark | GPT-5.5 | Grok 4.20 (Non-Reasoning) |
|---|---|---|
| LMArena Non-English | 1467 | 1440 |
| LMArena Chinese | 1535 | 1481 |
| LMArena French | 1486 | 1477 |
| LMArena German | 1478 | 1463 |
| LMArena Japanese | 1498 | 1450 |
| LMArena Korean | 1461 | 1417 |
| LMArena Russian | 1473 | 1457 |
| LMArena Spanish | 1469 | 1442 |

## Instruction Following

- GPT-5.5: 77.7 (#18)
- Grok 4.20 (Non-Reasoning): 75.0 (#79)

| Benchmark | GPT-5.5 | Grok 4.20 (Non-Reasoning) |
|---|---|---|
| LMArena Instruction Following | 1479 | 1420 |

## Long Context

- GPT-5.5: 48.3 (#12)
- Grok 4.20 (Non-Reasoning): 45.4 (#33)

| Benchmark | GPT-5.5 | Grok 4.20 (Non-Reasoning) |
|---|---|---|
| CL-bench Life | 22.2% | 11.9% |
| LMArena Longer Query | 1484 | 1436 |
| CL-bench | — | 22.2% |

## Writing & Preference

- GPT-5.5: 72.8 (#13)
- Grok 4.20 (Non-Reasoning): 65.9 (#42)

| Benchmark | GPT-5.5 | Grok 4.20 (Non-Reasoning) |
|---|---|---|
| LMArena Text | 1471 | 1451 |
| LMArena Creative Writing | 1455 | 1439 |
| EQ-Bench Creative Writing | 1844 | 1574 |
| LMArena Multi-Turn | 1477 | 1456 |
| EQ-Bench 4 | 1315 | — |

## FAQ

### Is GPT-5.5 better than Grok 4.20 (Non-Reasoning)?

GPT-5.5 is the stronger model overall, scoring 63.4 to 48.5 on the Noometry Index. Grok 4.20 (Non-Reasoning) costs 7.2× less per token, which makes it the better buy when GPT-5.5's lead doesn't matter for your workload.

### Which is cheaper, GPT-5.5 or Grok 4.20 (Non-Reasoning)?

Grok 4.20 (Non-Reasoning) is cheaper. It lists at $1.25 per million input tokens and $2.50 per million output tokens; GPT-5.5 lists at $5 and $30.

### Is GPT-5.5 or Grok 4.20 (Non-Reasoning) better for coding?

GPT-5.5 scores higher on coding benchmarks: 58.3 versus 41.8 in the Noometry coding category.

### Which has the bigger context window?

GPT-5.5 does, with 1.05M tokens against 1M.

### How many benchmarks do GPT-5.5 and Grok 4.20 (Non-Reasoning) share?

44 benchmarks have published results for both models. GPT-5.5 has 71 scored results on Noometry and Grok 4.20 (Non-Reasoning) has 46.