# GPT-6 Sol vs Wizardlm 70b

> GPT-6 Sol is the stronger model overall, scoring 61.8 to 33.0 on the Noometry Index.

- Canonical page: https://noometry.com/compare/gpt-6-sol-vs-wizardlm-70b
- Last updated: 2026-10-11
- Shared benchmarks: 12

## Summary

- They share 12 benchmarks with published results for both. GPT-6 Sol scores higher in 7 categories and Wizardlm 70b in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in math, where GPT-6 Sol leads 87.2 to 32.2.
- Wizardlm 70b has downloadable open weights; the other is API-only.

## Snapshot

| | GPT-6 Sol | Wizardlm 70b |
|---|---|---|
| Provider | OpenAI | Microsoft |
| Noometry Index | 61.8 | 33.0 |
| Rank | 12 | 249 |
| Context | 1.05M | — |
| Input $/M | $2 | — |
| Output $/M | $10 | — |
| Weights | Proprietary | Open |

## Coding

- GPT-6 Sol: 60.1 (#11)
- Wizardlm 70b: 31.4 (#285)

| Benchmark | GPT-6 Sol | Wizardlm 70b |
|---|---|---|
| LMArena Coding | 1447 | 1081 |
| DeepSWE | 68.8% | — |
| FrontierCode | 49.3% | — |
| LMArena WebDev | 1688 | — |
| SciCode | 57.6% | — |
| ALE-Bench | 2,462 | — |

## Agentic & Tool Use

- GPT-6 Sol: 37.2 (#36)
- Wizardlm 70b: —

| Benchmark | GPT-6 Sol | Wizardlm 70b |
|---|---|---|
| APEX-Agents | 54.3% | — |
| GDP.pdf | 26.4% | — |
| Vending-Bench 2 | 14,428 | — |

## Reasoning

- GPT-6 Sol: 74.0 (#9)
- Wizardlm 70b: 20.7 (#234)

| Benchmark | GPT-6 Sol | Wizardlm 70b |
|---|---|---|
| LMArena Hard Prompts | 1418 | 1079 |
| ARC-AGI-2 | 89.6% | — |
| NYT Connections (extended) | 90.1% | — |
| ARC-AGI-1 | 95.5% | — |
| CritPt | 30.9% | — |
| EBR-Bench | 53.3% | — |
| Mystery Game Puzzles | 56% | — |
| DTBench | 97.3% | — |
| LMCA | 59.1% | — |
| Epoch Capabilities Index | 162.72 | — |

## Math

- GPT-6 Sol: 87.2 (#7)
- Wizardlm 70b: 32.2 (#218)

| Benchmark | GPT-6 Sol | Wizardlm 70b |
|---|---|---|
| LMArena Math | 1402 | 1116 |
| FrontierMath (Tiers 1-3) | 89.8% | — |
| FrontierMath Tier 4 | 90% | — |
| OTIS Mock AIME 2024-2025 | 100% | — |
| ProofBench | 83% | — |

## Knowledge

- GPT-6 Sol: 64.8 (#15)
- Wizardlm 70b: —

| Benchmark | GPT-6 Sol | Wizardlm 70b |
|---|---|---|
| GPQA Diamond | 94.3% | — |
| SimpleQA Verified | 60.7% | — |
| Vectara Hallucination Rate | 6.5% | — |
| LMArena Expert | 1439 | — |

## Multimodal

- GPT-6 Sol: 47.6 (#10)
- Wizardlm 70b: —

| Benchmark | GPT-6 Sol | Wizardlm 70b |
|---|---|---|
| LMArena Vision | 1245 | — |
| Blueprint-Bench 2 | 36.9% | — |
| Furniture Assembly | 58.3% | — |

## Multilingual

- GPT-6 Sol: 50.5 (#118)
- Wizardlm 70b: 29.6 (#265)

| Benchmark | GPT-6 Sol | Wizardlm 70b |
|---|---|---|
| LMArena Non-English | 1385 | 1078 |
| LMArena Chinese | 1405 | 1052 |
| LMArena German | 1390 | 1083 |
| LMArena Russian | 1401 | 1155 |
| LMArena French | 1410 | — |
| LMArena Japanese | 1385 | — |
| LMArena Korean | 1341 | — |
| LMArena Spanish | 1384 | — |

## Instruction Following

- GPT-6 Sol: 74.5 (#94)
- Wizardlm 70b: 56.3 (#273)

| Benchmark | GPT-6 Sol | Wizardlm 70b |
|---|---|---|
| LMArena Instruction Following | 1412 | 1093 |

## Long Context

- GPT-6 Sol: 43.1 (#108)
- Wizardlm 70b: 33.3 (#263)

| Benchmark | GPT-6 Sol | Wizardlm 70b |
|---|---|---|
| LMArena Longer Query | 1411 | 1097 |

## Writing & Preference

- GPT-6 Sol: 71.9 (#18)
- Wizardlm 70b: 34.8 (#269)

| Benchmark | GPT-6 Sol | Wizardlm 70b |
|---|---|---|
| LMArena Text | 1395 | 1120 |
| LMArena Creative Writing | 1378 | 1149 |
| LMArena Multi-Turn | 1412 | 1108 |
| EQ-Bench Creative Writing | 2125 | — |

## FAQ

### Is GPT-6 Sol better than Wizardlm 70b?

GPT-6 Sol is the stronger model overall, scoring 61.8 to 33.0 on the Noometry Index.

### Is GPT-6 Sol or Wizardlm 70b better for coding?

GPT-6 Sol scores higher on coding benchmarks: 60.1 versus 31.4 in the Noometry coding category.

### How many benchmarks do GPT-6 Sol and Wizardlm 70b share?

12 benchmarks have published results for both models. GPT-6 Sol has 45 scored results on Noometry and Wizardlm 70b has 12.