Model comparison
Grok 4.3 vs Grok 4.6
Grok 4.6 is the stronger model overall, scoring 56.8 to 43.6 on the Noometry Index. Grok 4.3 costs 1.9× less per token, which makes it the better buy when Grok 4.6's lead doesn't matter for your workload.
Last verified . 37 shared benchmarks.
Summary
- They share 37 benchmarks with published results for both. Grok 4.3 scores higher in 0 categories and Grok 4.6 in 10 categories; 10 gaps are clear of the uncertainty.
- The widest gap is in reasoning, where Grok 4.6 leads 61.3 to 34.9.
- The biggest single-benchmark swing is ProofBench: 11% for Grok 4.3 and 51% for Grok 4.6.
- Grok 4.3 is cheaper at $1.25 / $2.50 per million input/output tokens, against $2 / $6 for Grok 4.6.
- Grok 4.3 accepts more context: 1M tokens versus 500K.
Side by side
| Grok 4.3 | Grok 4.6 | |
|---|---|---|
| Provider | xAI | xAI |
| Noometry Index | 43.6 | 56.8 |
| Released | 2026-04-17 | 2026-08-12 |
| Weights | Proprietary | Proprietary |
| Context window | 1M | 500K |
| Max output | 30K | 500K |
| Input $ / M tokens | $1.25 | $2 |
| Output $ / M tokens | $2.50 | $6 |
| Results tracked | 40 | 49 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4.6 leads
Grok 4.3: 41.3 (#117), Grok 4.6: 58.0 (#14)
| Benchmark | Grok 4.3 | Grok 4.6 |
|---|---|---|
| LMArena WebDev | 1357 | 1620 |
| SciCode | 47.3% | 56.5% |
| WeirdML | 49.9% | 67.3% |
| LMArena Coding | 1415 | 1464 |
| ALE-Bench | 944.17 | 1,508 |
| DeepSWE | — | 67.5% |
| FrontierCode | — | 48% |
| CursorBench | — | 41.4% |
| FrontierSWE | — | 25.3% |
Agentic & Tool Use Grok 4.6 leads
Grok 4.3: 27.6 (#93), Grok 4.6: 39.4 (#25)
| Benchmark | Grok 4.3 | Grok 4.6 |
|---|---|---|
| GDP.pdf | 8% | 17.2% |
| Vending-Bench 2 | 35.26 | 9,047 |
| APEX-Agents | — | 65.3% |
| LMArena Search | 1165 | — |
Reasoning Grok 4.6 leads
Grok 4.3: 34.9 (#60), Grok 4.6: 61.3 (#18)
| Benchmark | Grok 4.3 | Grok 4.6 |
|---|---|---|
| NYT Connections (extended) | 55.2% | 80% |
| CritPt | 8% | 19.7% |
| Chess Puzzles | 25% | 40% |
| LMArena Hard Prompts | 1397 | 1445 |
| DTBench | 90.7% | 97.3% |
| LMCA | 38.3% | 48.5% |
| Epoch Capabilities Index | 149.13 | 156.48 |
| ARC-AGI-2 | — | 67.1% |
| SimpleBench | — | 75.9% |
| ARC-AGI-1 | — | 87.5% |
| EBR-Bench | — | 30.5% |
| Mystery Game Puzzles | — | 34% |
| ForecastBench | 60.3 | — |
Math Grok 4.6 leads
Grok 4.3: 45.8 (#60), Grok 4.6: 66.8 (#19)
| Benchmark | Grok 4.3 | Grok 4.6 |
|---|---|---|
| FrontierMath (Tiers 1-3) | 42.8% | 66% |
| FrontierMath Tier 4 | 14.6% | 31.7% |
| OTIS Mock AIME 2024-2025 | 93.3% | 99.2% |
| ProofBench | 11% | 51% |
| LMArena Math | 1388 | 1416 |
Knowledge Grok 4.6 leads
Grok 4.3: 52.5 (#57), Grok 4.6: 63.3 (#18)
| Benchmark | Grok 4.3 | Grok 4.6 |
|---|---|---|
| GPQA Diamond | 88.8% | 94% |
| SimpleQA Verified | 33.2% | 49.3% |
| LMArena Expert | 1385 | 1463 |
Multimodal Grok 4.6 leads
Grok 4.3: 31.6 (#96), Grok 4.6: 43.9 (#18)
| Benchmark | Grok 4.3 | Grok 4.6 |
|---|---|---|
| LMArena Vision | 1231 | 1265 |
| Blueprint-Bench 2 | 0% | 33.2% |
| Furniture Assembly | — | 40% |
| LMArena Document | — | 1452 |
Multilingual Grok 4.6 leads
Grok 4.3: 50.7 (#114), Grok 4.6: 53.3 (#67)
| Benchmark | Grok 4.3 | Grok 4.6 |
|---|---|---|
| LMArena Non-English | 1385 | 1422 |
| LMArena Chinese | 1423 | 1479 |
| LMArena French | 1412 | 1458 |
| LMArena German | 1395 | 1425 |
| LMArena Japanese | 1379 | 1383 |
| LMArena Korean | 1357 | 1395 |
| LMArena Russian | 1399 | 1427 |
| LMArena Spanish | 1402 | 1408 |
Instruction Following Grok 4.6 leads
Grok 4.3: 72.3 (#135), Grok 4.6: 75.6 (#59)
| Benchmark | Grok 4.3 | Grok 4.6 |
|---|---|---|
| LMArena Instruction Following | 1366 | 1431 |
Long Context Grok 4.6 leads
Grok 4.3: 42.5 (#118), Grok 4.6: 44.4 (#62)
| Benchmark | Grok 4.3 | Grok 4.6 |
|---|---|---|
| LMArena Longer Query | 1393 | 1453 |
Writing & Preference Grok 4.6 leads
Grok 4.3: 58.6 (#112), Grok 4.6: 62.5 (#77)
| Benchmark | Grok 4.3 | Grok 4.6 |
|---|---|---|
| LMArena Text | 1397 | 1427 |
| LMArena Creative Writing | 1380 | 1430 |
| LMArena Multi-Turn | 1406 | 1422 |
| EQ-Bench 4 | 1075 | — |
Frequently asked questions
Is Grok 4.3 better than Grok 4.6?
Grok 4.6 is the stronger model overall, scoring 56.8 to 43.6 on the Noometry Index. Grok 4.3 costs 1.9× less per token, which makes it the better buy when Grok 4.6's lead doesn't matter for your workload.
Which is cheaper, Grok 4.3 or Grok 4.6?
Grok 4.3 is cheaper. It lists at $1.25 per million input tokens and $2.50 per million output tokens; Grok 4.6 lists at $2 and $6.
Is Grok 4.3 or Grok 4.6 better for coding?
Grok 4.6 scores higher on coding benchmarks: 58.0 versus 41.3 in the Noometry coding category.
Which has the bigger context window?
Grok 4.3 does, with 1M tokens against 500K.
How many benchmarks do Grok 4.3 and Grok 4.6 share?
37 benchmarks have published results for both models. Grok 4.3 has 40 scored results on Noometry and Grok 4.6 has 49.