Model comparison
Grok-2 (Dec 2024) vs Olmo 3 32b Think
Olmo 3 32b Think is the stronger model overall, scoring 38.7 to 33.7 on the Noometry Index.
Last verified . 14 shared benchmarks.
Allen Institute for AI (Ai2)Summary
- They share 14 benchmarks with published results for both. Grok-2 (Dec 2024) scores higher in 1 category and Olmo 3 32b Think in 7 categories; 5 gaps are clear of the uncertainty.
- The widest gap is in math, where Olmo 3 32b Think leads 36.5 to 20.8.
- Olmo 3 32b Think has downloadable open weights; the other is API-only.
Side by side
| Grok-2 (Dec 2024) | Olmo 3 32b Think | |
|---|---|---|
| Provider | xAI | Allen Institute for AI (Ai2) |
| Noometry Index | 33.7 | 38.7 |
| Released | 2024-08-13 | — |
| Weights | Proprietary | Open |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 34 | 14 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Olmo 3 32b Think leads
Grok-2 (Dec 2024): 33.3 (#258), Olmo 3 32b Think: 38.6 (#172)
| Benchmark | Grok-2 (Dec 2024) | Olmo 3 32b Think |
|---|---|---|
| LMArena Coding | 1287 | 1319 |
| WeirdML | 22.2% | — |
| LiveBench Coding | 46.4% | — |
Reasoning Olmo 3 32b Think leads
Grok-2 (Dec 2024): 16.9 (#299), Olmo 3 32b Think: 25.9 (#140)
| Benchmark | Grok-2 (Dec 2024) | Olmo 3 32b Think |
|---|---|---|
| LMArena Hard Prompts | 1272 | 1302 |
| SimpleBench | 22.7% | — |
| LiveBench Reasoning | 54.8% | — |
| DTBench | 65.2% | — |
| LiveBench Data Analysis | 54.5% | — |
| Epoch Capabilities Index | 130.48 | — |
| LiveBench | 54.3% | — |
Math Olmo 3 32b Think leads
Grok-2 (Dec 2024): 20.8 (#284), Olmo 3 32b Think: 36.5 (#165)
| Benchmark | Grok-2 (Dec 2024) | Olmo 3 32b Think |
|---|---|---|
| LMArena Math | 1283 | 1316 |
| OTIS Mock AIME 2024-2025 | 11.5% | — |
| LiveBench Math | 54.9% | — |
| MATH Level 5 | 63.5% | — |
| FrontierMath (Feb 2025 set) | 0.7% | — |
Knowledge Olmo 3 32b Think leads
Grok-2 (Dec 2024): 29.8 (#233), Olmo 3 32b Think: 35.0 (#190)
| Benchmark | Grok-2 (Dec 2024) | Olmo 3 32b Think |
|---|---|---|
| LMArena Expert | 1254 | 1273 |
| GPQA Diamond | 53.8% | — |
| Confabulations | 20.1% | — |
Multilingual Grok-2 (Dec 2024) leads
Grok-2 (Dec 2024): 43.1 (#188), Olmo 3 32b Think: 41.2 (#210)
| Benchmark | Grok-2 (Dec 2024) | Olmo 3 32b Think |
|---|---|---|
| LMArena Non-English | 1282 | 1255 |
| LMArena Chinese | 1289 | 1300 |
| LMArena French | 1318 | 1291 |
| LMArena German | 1287 | 1290 |
| LMArena Russian | 1286 | 1254 |
| LMArena Japanese | 1244 | — |
| LMArena Korean | 1237 | — |
| LMArena Spanish | 1281 | — |
Instruction Following Too close to call
Grok-2 (Dec 2024): 66.9 (#202), Olmo 3 32b Think: 67.2 (#198)
| Benchmark | Grok-2 (Dec 2024) | Olmo 3 32b Think |
|---|---|---|
| LMArena Instruction Following | 1270 | 1275 |
| LiveBench Instruction Following | 69.6% | — |
Long Context Too close to call
Grok-2 (Dec 2024): 38.8 (#190), Olmo 3 32b Think: 39.4 (#182)
| Benchmark | Grok-2 (Dec 2024) | Olmo 3 32b Think |
|---|---|---|
| LMArena Longer Query | 1276 | 1296 |
Writing & Preference Too close to call
Grok-2 (Dec 2024): 48.6 (#198), Olmo 3 32b Think: 49.1 (#193)
| Benchmark | Grok-2 (Dec 2024) | Olmo 3 32b Think |
|---|---|---|
| LMArena Text | 1305 | 1300 |
| LMArena Creative Writing | 1284 | 1256 |
| LMArena Multi-Turn | 1290 | 1290 |
| Short-Story Creative Writing | 63.6% | — |
| LiveBench Language | 45.6% | — |
Frequently asked questions
Is Grok-2 (Dec 2024) better than Olmo 3 32b Think?
Olmo 3 32b Think is the stronger model overall, scoring 38.7 to 33.7 on the Noometry Index.
Is Grok-2 (Dec 2024) or Olmo 3 32b Think better for coding?
Olmo 3 32b Think scores higher on coding benchmarks: 38.6 versus 33.3 in the Noometry coding category.
How many benchmarks do Grok-2 (Dec 2024) and Olmo 3 32b Think share?
14 benchmarks have published results for both models. Grok-2 (Dec 2024) has 34 scored results on Noometry and Olmo 3 32b Think has 14.