# Grok 4.7 vs Hy3

> Grok 4.7 is the stronger model overall, scoring 53.1 to 44.2 on the Noometry Index. Hy3 costs 13× less per token, which makes it the better buy when Grok 4.7's lead doesn't matter for your workload.

- Canonical page: https://noometry.com/compare/grok-4-7-vs-hy3
- Last updated: 2026-10-11
- Shared benchmarks: 16

## Summary

- They share 16 benchmarks with published results for both. Grok 4.7 scores higher in 5 categories and Hy3 in 3 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Grok 4.7 leads 49.1 to 26.1.
- The biggest single-benchmark swing is NYT Connections (extended): 76.8% for Grok 4.7 and 41.2% for Hy3.
- Hy3 is cheaper at $0.13 / $0.53 per million input/output tokens, against $2 / $6 for Grok 4.7.
- Grok 4.7 accepts more context: 500K tokens versus 262K.
- Hy3 has downloadable open weights; the other is API-only.

## Snapshot

| | Grok 4.7 | Hy3 |
|---|---|---|
| Provider | xAI | Tencent |
| Noometry Index | 53.1 | 44.2 |
| Rank | 37 | 79 |
| Context | 500K | 262K |
| Input $/M | $2 | $0.13 |
| Output $/M | $6 | $0.53 |
| Weights | Proprietary | Open |

## Coding

- Grok 4.7: 58.0 (#18)
- Hy3: 46.8 (#63)

| Benchmark | Grok 4.7 | Hy3 |
|---|---|---|
| LMArena WebDev | 1639 | 1508 |
| LMArena Coding | 1427 | 1464 |
| FrontierCode | 47.6% | — |
| CursorBench | 46.3% | — |
| FrontierSWE | 29.5% | — |
| SciCode | 57.8% | — |

## Agentic & Tool Use

- Grok 4.7: 36.7 (#37)
- Hy3: —

| Benchmark | Grok 4.7 | Hy3 |
|---|---|---|
| APEX-Agents | 54.6% | — |
| GDP.pdf | 22.8% | — |
| Vending-Bench 2 | 10,537 | — |

## Reasoning

- Grok 4.7: 49.1 (#40)
- Hy3: 26.1 (#136)

| Benchmark | Grok 4.7 | Hy3 |
|---|---|---|
| NYT Connections (extended) | 76.8% | 41.2% |
| LMArena Hard Prompts | 1413 | 1447 |
| CritPt | 18% | — |
| Chess Puzzles | 38% | — |
| Mystery Game Puzzles | 29% | — |
| DTBench | 96% | — |
| LMCA | 49.4% | — |
| Epoch Capabilities Index | 153.53 | — |

## Math

- Grok 4.7: 57.8 (#39)
- Hy3: 40.1 (#93)

| Benchmark | Grok 4.7 | Hy3 |
|---|---|---|
| LMArena Math | 1407 | 1475 |
| FrontierMath (Tiers 1-3) | 53% | — |
| FrontierMath Tier 4 | 17.1% | — |
| OTIS Mock AIME 2024-2025 | 98.1% | — |
| ProofBench | 34% | — |

## Knowledge

- Grok 4.7: 62.8 (#22)
- Hy3: 40.8 (#114)

| Benchmark | Grok 4.7 | Hy3 |
|---|---|---|
| LMArena Expert | 1422 | 1460 |
| GPQA Diamond | 92.7% | — |
| SimpleQA Verified | 56% | — |

## Multimodal

- Grok 4.7: 35.5 (#87)
- Hy3: —

| Benchmark | Grok 4.7 | Hy3 |
|---|---|---|
| LMArena Vision | 1228 | — |
| Blueprint-Bench 2 | 32.5% | — |
| Furniture Assembly | 20.8% | — |

## Multilingual

- Grok 4.7: 50.8 (#116)
- Hy3: 53.5 (#65)

| Benchmark | Grok 4.7 | Hy3 |
|---|---|---|
| LMArena Non-English | 1389 | 1426 |
| LMArena Chinese | 1455 | 1493 |
| LMArena French | 1455 | 1461 |
| LMArena Russian | 1397 | 1432 |
| LMArena Spanish | 1400 | 1456 |
| LMArena German | — | 1439 |
| LMArena Japanese | — | 1392 |
| LMArena Korean | — | 1395 |

## Instruction Following

- Grok 4.7: 74.1 (#105)
- Hy3: 75.1 (#70)

| Benchmark | Grok 4.7 | Hy3 |
|---|---|---|
| LMArena Instruction Following | 1404 | 1426 |

## Long Context

- Grok 4.7: 43.1 (#104)
- Hy3: 44.1 (#75)

| Benchmark | Grok 4.7 | Hy3 |
|---|---|---|
| LMArena Longer Query | 1413 | 1442 |

## Writing & Preference

- Grok 4.7: 70.0 (#24)
- Hy3: 62.2 (#81)

| Benchmark | Grok 4.7 | Hy3 |
|---|---|---|
| LMArena Text | 1399 | 1439 |
| LMArena Creative Writing | 1391 | 1402 |
| LMArena Multi-Turn | 1393 | 1436 |
| EQ-Bench Creative Writing | 2007 | — |

## FAQ

### Is Grok 4.7 better than Hy3?

Grok 4.7 is the stronger model overall, scoring 53.1 to 44.2 on the Noometry Index. Hy3 costs 13× less per token, which makes it the better buy when Grok 4.7's lead doesn't matter for your workload.

### Which is cheaper, Grok 4.7 or Hy3?

Hy3 is cheaper. It lists at $0.13 per million input tokens and $0.53 per million output tokens; Grok 4.7 lists at $2 and $6.

### Is Grok 4.7 or Hy3 better for coding?

Grok 4.7 scores higher on coding benchmarks: 58.0 versus 46.8 in the Noometry coding category.

### Which has the bigger context window?

Grok 4.7 does, with 500K tokens against 262K.

### How many benchmarks do Grok 4.7 and Hy3 share?

16 benchmarks have published results for both models. Grok 4.7 has 39 scored results on Noometry and Hy3 has 19.