Model comparison
Grok 3 vs Inkling
Inkling is the stronger model overall, scoring 44.1 to 39.9 on the Noometry Index.
Last verified . 25 shared benchmarks.
Summary
- They share 25 benchmarks with published results for both. Grok 3 scores higher in 3 categories and Inkling in 6 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Inkling leads 40.4 to 13.7.
- The biggest single-benchmark swing is ARC-AGI-1: 5.5% for Grok 3 and 79.5% for Inkling.
- Inkling has downloadable open weights; the other is API-only.
Side by side
| Grok 3 | Inkling | |
|---|---|---|
| Provider | xAI | Thinking Machines Lab |
| Noometry Index | 39.9 | 44.1 |
| Released | 2025-04-09 | 2026-07-15 |
| Weights | Proprietary | Open |
| Context window | — | 66K |
| Max output | — | 66K |
| Input $ / M tokens | — | $1.87 |
| Output $ / M tokens | — | $4.68 |
| Results tracked | 40 | 41 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 3 leads
Grok 3: 41.9 (#115), Inkling: 34.5 (#234)
| Benchmark | Grok 3 | Inkling |
|---|---|---|
| WeirdML | 37.2% | 32.3% |
| LMArena Coding | 1432 | 1464 |
| FrontierCode | — | 14% |
| Aider Polyglot | 53.3% | — |
| LMArena WebDev | — | 1413 |
| FrontierSWE | — | 4.1% |
| SciCode | — | 47% |
| ALE-Bench | — | 946 |
Agentic & Tool Use Too close to call
Grok 3: 30.5 (#76), Inkling: 29.6 (#85)
| Benchmark | Grok 3 | Inkling |
|---|---|---|
| APEX-Agents | — | 33.8% |
| τ²-bench Banking | — | 25% |
| BALROG | 29.5% | — |
Reasoning Inkling leads
Grok 3: 13.7 (#333), Inkling: 40.4 (#56)
| Benchmark | Grok 3 | Inkling |
|---|---|---|
| ARC-AGI-2 | 0% | 36.5% |
| SimpleBench | 36.1% | 50% |
| ARC-AGI-1 | 5.5% | 79.5% |
| LMArena Hard Prompts | 1434 | 1451 |
| Epoch Capabilities Index | 138.33 | 148.54 |
| Kagi LLM Benchmark | 61.3% | — |
| CritPt | — | 5.4% |
| Chess Puzzles | — | 21% |
| DTBench | — | 87.5% |
| LMCA | — | 37.6% |
Math Grok 3 leads
Grok 3: 38.0 (#145), Inkling: 31.3 (#225)
| Benchmark | Grok 3 | Inkling |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 55.6% | 88.9% |
| LMArena Math | 1391 | 1479 |
| FrontierMath (Tiers 1-3) | — | 33.3% |
| FrontierMath Tier 4 | — | 4.9% |
| ProofBench | — | 0% |
| Omni-MATH | 46.4% | — |
| MATH Level 5 | 88.7% | — |
| FrontierMath (Feb 2025 set) | 3.8% | — |
| FrontierMath Tier 4 (v1) | 0% | — |
Knowledge Inkling leads
Grok 3: 46.2 (#82), Inkling: 55.1 (#49)
| Benchmark | Grok 3 | Inkling |
|---|---|---|
| GPQA Diamond | 75.8% | 88.3% |
| LMArena Expert | 1421 | 1465 |
| SimpleQA Verified | — | 40.3% |
| MMLU-Pro | 78.8% | — |
| Confabulations | 14.2% | — |
| Vectara Hallucination Rate | 5.8% | — |
| GPQA (HELM) | 65% | — |
Multilingual Inkling leads
Grok 3: 52.3 (#87), Inkling: 54.0 (#52)
| Benchmark | Grok 3 | Inkling |
|---|---|---|
| LMArena Non-English | 1410 | 1434 |
| LMArena Chinese | 1448 | 1490 |
| LMArena French | 1460 | 1458 |
| LMArena German | 1431 | 1446 |
| LMArena Japanese | 1387 | 1429 |
| LMArena Korean | 1373 | 1404 |
| LMArena Russian | 1416 | 1429 |
| LMArena Spanish | 1417 | 1448 |
Instruction Following Too close to call
Grok 3: 75.0 (#73), Inkling: 75.1 (#71)
| Benchmark | Grok 3 | Inkling |
|---|---|---|
| LMArena Instruction Following | 1409 | 1426 |
| IFEval | 88.4% | — |
Long Context Inkling leads
Grok 3: 38.7 (#192), Inkling: 43.8 (#86)
| Benchmark | Grok 3 | Inkling |
|---|---|---|
| LMArena Longer Query | 1439 | 1434 |
| Fiction.LiveBench | 58.3% | — |
Writing & Preference Inkling leads
Grok 3: 55.8 (#141), Inkling: 65.2 (#51)
| Benchmark | Grok 3 | Inkling |
|---|---|---|
| LMArena Text | 1426 | 1441 |
| LMArena Creative Writing | 1414 | 1387 |
| EQ-Bench Creative Writing | 1186 | 1611 |
| LMArena Multi-Turn | 1425 | 1436 |
| Short-Story Creative Writing | 76.4% | — |
| WildBench | 84.9% | — |
| EQ-Bench 4 | — | 1226 |
Frequently asked questions
Is Grok 3 better than Inkling?
Inkling is the stronger model overall, scoring 44.1 to 39.9 on the Noometry Index.
Is Grok 3 or Inkling better for coding?
Grok 3 scores higher on coding benchmarks: 41.9 versus 34.5 in the Noometry coding category.
How many benchmarks do Grok 3 and Inkling share?
25 benchmarks have published results for both models. Grok 3 has 40 scored results on Noometry and Inkling has 41.