Model comparison
Granite 4.2 8B vs Grok 4.3
Grok 4.3 is the stronger model overall, scoring 43.6 to 40.0 on the Noometry Index. Granite 4.2 8B costs 15× less per token, which makes it the better buy when Grok 4.3's lead doesn't matter for your workload.
Last verified . 11 shared benchmarks.
Summary
- They share 11 benchmarks with published results for both. Granite 4.2 8B scores higher in 0 categories and Grok 4.3 in 7 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Grok 4.3 leads 52.5 to 37.7.
- Granite 4.2 8B is cheaper at $0.06 / $0.25 per million input/output tokens, against $1.25 / $2.50 for Grok 4.3.
- Grok 4.3 accepts more context: 1M tokens versus 131K.
- Granite 4.2 8B has downloadable open weights; the other is API-only.
Side by side
| Granite 4.2 8B | Grok 4.3 | |
|---|---|---|
| Provider | IBM | xAI |
| Noometry Index | 40.0 | 43.6 |
| Released | — | 2026-04-17 |
| Weights | Open | Proprietary |
| Context window | 131K | 1M |
| Max output | 118K | 30K |
| Input $ / M tokens | $0.06 | $1.25 |
| Output $ / M tokens | $0.25 | $2.50 |
| Results tracked | 11 | 40 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4.3 leads
Granite 4.2 8B: 39.9 (#138), Grok 4.3: 41.3 (#117)
| Benchmark | Granite 4.2 8B | Grok 4.3 |
|---|---|---|
| LMArena Coding | 1379 | 1415 |
| LMArena WebDev | — | 1357 |
| SciCode | — | 47.3% |
| WeirdML | — | 49.9% |
| ALE-Bench | — | 944.17 |
Agentic & Tool Use Not comparable
Granite 4.2 8B: —, Grok 4.3: 27.6 (#93)
| Benchmark | Granite 4.2 8B | Grok 4.3 |
|---|---|---|
| GDP.pdf | — | 8% |
| LMArena Search | — | 1165 |
| Vending-Bench 2 | — | 35.26 |
Reasoning Grok 4.3 leads
Granite 4.2 8B: 25.5 (#130), Grok 4.3: 34.9 (#60)
| Benchmark | Granite 4.2 8B | Grok 4.3 |
|---|---|---|
| LMArena Hard Prompts | 1327 | 1397 |
| NYT Connections (extended) | — | 55.2% |
| CritPt | — | 8% |
| Chess Puzzles | — | 25% |
| DTBench | — | 90.7% |
| LMCA | — | 38.3% |
| Epoch Capabilities Index | — | 149.13 |
| ForecastBench | — | 60.3 |
Math Not comparable
Granite 4.2 8B: —, Grok 4.3: 45.8 (#60)
| Benchmark | Granite 4.2 8B | Grok 4.3 |
|---|---|---|
| FrontierMath (Tiers 1-3) | — | 42.8% |
| FrontierMath Tier 4 | — | 14.6% |
| OTIS Mock AIME 2024-2025 | — | 93.3% |
| ProofBench | — | 11% |
| LMArena Math | — | 1388 |
Knowledge Grok 4.3 leads
Granite 4.2 8B: 37.7 (#144), Grok 4.3: 52.5 (#57)
| Benchmark | Granite 4.2 8B | Grok 4.3 |
|---|---|---|
| LMArena Expert | 1380 | 1385 |
| GPQA Diamond | — | 88.8% |
| SimpleQA Verified | — | 33.2% |
Multimodal Not comparable
Granite 4.2 8B: —, Grok 4.3: 31.6 (#96)
| Benchmark | Granite 4.2 8B | Grok 4.3 |
|---|---|---|
| LMArena Vision | — | 1231 |
| Blueprint-Bench 2 | — | 0% |
Multilingual Grok 4.3 leads
Granite 4.2 8B: 44.5 (#173), Grok 4.3: 50.7 (#114)
| Benchmark | Granite 4.2 8B | Grok 4.3 |
|---|---|---|
| LMArena Non-English | 1299 | 1385 |
| LMArena Chinese | 1361 | 1423 |
| LMArena Russian | 1283 | 1399 |
| LMArena French | — | 1412 |
| LMArena German | — | 1395 |
| LMArena Japanese | — | 1379 |
| LMArena Korean | — | 1357 |
| LMArena Spanish | — | 1402 |
Instruction Following Grok 4.3 leads
Granite 4.2 8B: 68.7 (#183), Grok 4.3: 72.3 (#135)
| Benchmark | Granite 4.2 8B | Grok 4.3 |
|---|---|---|
| LMArena Instruction Following | 1298 | 1366 |
Long Context Grok 4.3 leads
Granite 4.2 8B: 40.2 (#154), Grok 4.3: 42.5 (#118)
| Benchmark | Granite 4.2 8B | Grok 4.3 |
|---|---|---|
| LMArena Longer Query | 1323 | 1393 |
Writing & Preference Grok 4.3 leads
Granite 4.2 8B: 49.8 (#183), Grok 4.3: 58.6 (#112)
| Benchmark | Granite 4.2 8B | Grok 4.3 |
|---|---|---|
| LMArena Text | 1319 | 1397 |
| LMArena Creative Writing | 1238 | 1380 |
| LMArena Multi-Turn | 1299 | 1406 |
| EQ-Bench 4 | — | 1075 |
Frequently asked questions
Is Granite 4.2 8B better than Grok 4.3?
Grok 4.3 is the stronger model overall, scoring 43.6 to 40.0 on the Noometry Index. Granite 4.2 8B costs 15× less per token, which makes it the better buy when Grok 4.3's lead doesn't matter for your workload.
Which is cheaper, Granite 4.2 8B or Grok 4.3?
Granite 4.2 8B is cheaper. It lists at $0.06 per million input tokens and $0.25 per million output tokens; Grok 4.3 lists at $1.25 and $2.50.
Is Granite 4.2 8B or Grok 4.3 better for coding?
Grok 4.3 scores higher on coding benchmarks: 41.3 versus 39.9 in the Noometry coding category.
Which has the bigger context window?
Grok 4.3 does, with 1M tokens against 131K.
How many benchmarks do Granite 4.2 8B and Grok 4.3 share?
11 benchmarks have published results for both models. Granite 4.2 8B has 11 scored results on Noometry and Grok 4.3 has 40.