Model comparison
Granite 3.1 8b Instruct vs Grok 3
Grok 3 is the stronger model overall, scoring 39.9 to 32.4 on the Noometry Index.
Last verified . 12 shared benchmarks.
Summary
- They share 12 benchmarks with published results for both. Granite 3.1 8b Instruct scores higher in 1 category and Grok 3 in 8 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in multilingual, where Grok 3 leads 52.3 to 30.9.
- Granite 3.1 8b Instruct has downloadable open weights; the other is API-only.
Side by side
| Granite 3.1 8b Instruct | Grok 3 | |
|---|---|---|
| Provider | IBM | xAI |
| Noometry Index | 32.4 | 39.9 |
| Released | — | 2025-04-09 |
| Weights | Open | Proprietary |
| Context window | — | — |
| Max output | — | — |
| Input $ / M tokens | — | — |
| Output $ / M tokens | — | — |
| Results tracked | 13 | 40 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 3 leads
Granite 3.1 8b Instruct: 34.5 (#233), Grok 3: 41.9 (#115)
| Benchmark | Granite 3.1 8b Instruct | Grok 3 |
|---|---|---|
| LMArena Coding | 1186 | 1432 |
| Aider Polyglot | — | 53.3% |
| WeirdML | — | 37.2% |
Agentic & Tool Use Grok 3 leads
Granite 3.1 8b Instruct: 24.1 (#120), Grok 3: 30.5 (#76)
| Benchmark | Granite 3.1 8b Instruct | Grok 3 |
|---|---|---|
| Berkeley Function Calling Leaderboard | 27.1% | — |
| BALROG | — | 29.5% |
Reasoning Granite 3.1 8b Instruct leads
Granite 3.1 8b Instruct: 22.1 (#207), Grok 3: 13.7 (#333)
| Benchmark | Granite 3.1 8b Instruct | Grok 3 |
|---|---|---|
| LMArena Hard Prompts | 1145 | 1434 |
| ARC-AGI-2 | — | 0% |
| SimpleBench | — | 36.1% |
| Kagi LLM Benchmark | — | 61.3% |
| ARC-AGI-1 | — | 5.5% |
| Epoch Capabilities Index | — | 138.33 |
Math Grok 3 leads
Granite 3.1 8b Instruct: 33.0 (#209), Grok 3: 38.0 (#145)
| Benchmark | Granite 3.1 8b Instruct | Grok 3 |
|---|---|---|
| LMArena Math | 1152 | 1391 |
| OTIS Mock AIME 2024-2025 | — | 55.6% |
| Omni-MATH | — | 46.4% |
| MATH Level 5 | — | 88.7% |
| FrontierMath (Feb 2025 set) | — | 3.8% |
| FrontierMath Tier 4 (v1) | — | 0% |
Knowledge Grok 3 leads
Granite 3.1 8b Instruct: 31.1 (#220), Grok 3: 46.2 (#82)
| Benchmark | Granite 3.1 8b Instruct | Grok 3 |
|---|---|---|
| LMArena Expert | 1142 | 1421 |
| GPQA Diamond | — | 75.8% |
| MMLU-Pro | — | 78.8% |
| Confabulations | — | 14.2% |
| Vectara Hallucination Rate | — | 5.8% |
| GPQA (HELM) | — | 65% |
Multilingual Grok 3 leads
Granite 3.1 8b Instruct: 30.9 (#260), Grok 3: 52.3 (#87)
| Benchmark | Granite 3.1 8b Instruct | Grok 3 |
|---|---|---|
| LMArena Non-English | 1099 | 1410 |
| LMArena Chinese | 1145 | 1448 |
| LMArena Russian | 1092 | 1416 |
| LMArena French | — | 1460 |
| LMArena German | — | 1431 |
| LMArena Japanese | — | 1387 |
| LMArena Korean | — | 1373 |
| LMArena Spanish | — | 1417 |
Instruction Following Grok 3 leads
Granite 3.1 8b Instruct: 58.6 (#259), Grok 3: 75.0 (#73)
| Benchmark | Granite 3.1 8b Instruct | Grok 3 |
|---|---|---|
| LMArena Instruction Following | 1131 | 1409 |
| IFEval | — | 88.4% |
Long Context Grok 3 leads
Granite 3.1 8b Instruct: 35.2 (#241), Grok 3: 38.7 (#192)
| Benchmark | Granite 3.1 8b Instruct | Grok 3 |
|---|---|---|
| LMArena Longer Query | 1162 | 1439 |
| Fiction.LiveBench | — | 58.3% |
Writing & Preference Grok 3 leads
Granite 3.1 8b Instruct: 35.5 (#266), Grok 3: 55.8 (#141)
| Benchmark | Granite 3.1 8b Instruct | Grok 3 |
|---|---|---|
| LMArena Text | 1150 | 1426 |
| LMArena Creative Writing | 1129 | 1414 |
| LMArena Multi-Turn | 1108 | 1425 |
| Short-Story Creative Writing | — | 76.4% |
| EQ-Bench Creative Writing | — | 1186 |
| WildBench | — | 84.9% |
Frequently asked questions
Is Granite 3.1 8b Instruct better than Grok 3?
Grok 3 is the stronger model overall, scoring 39.9 to 32.4 on the Noometry Index.
Is Granite 3.1 8b Instruct or Grok 3 better for coding?
Grok 3 scores higher on coding benchmarks: 41.9 versus 34.5 in the Noometry coding category.
How many benchmarks do Granite 3.1 8b Instruct and Grok 3 share?
12 benchmarks have published results for both models. Granite 3.1 8b Instruct has 13 scored results on Noometry and Grok 3 has 40.