Model comparison
Grok 4.3 vs Llama2 70b Steerlm Chat
Grok 4.3 is the stronger model overall, scoring 43.6 to 31.2 on the Noometry Index.
Last verified . 9 shared benchmarks.
NVIDIASummary
- They share 9 benchmarks with published results for both. Grok 4.3 scores higher in 7 categories and Llama2 70b Steerlm Chat in 0 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Grok 4.3 leads 58.6 to 31.8.
- Llama2 70b Steerlm Chat has downloadable open weights; the other is API-only.
Side by side
| Grok 4.3 | Llama2 70b Steerlm Chat | |
|---|---|---|
| Provider | xAI | NVIDIA |
| Noometry Index | 43.6 | 31.2 |
| Released | 2026-04-17 | — |
| Weights | Proprietary | Open |
| Context window | 1M | — |
| Max output | 30K | — |
| Input $ / M tokens | $1.25 | — |
| Output $ / M tokens | $2.50 | — |
| Results tracked | 40 | 9 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4.3 leads
Grok 4.3: 41.3 (#117), Llama2 70b Steerlm Chat: 29.3 (#295)
| Benchmark | Grok 4.3 | Llama2 70b Steerlm Chat |
|---|---|---|
| LMArena Coding | 1415 | 1025 |
| LMArena WebDev | 1357 | — |
| SciCode | 47.3% | — |
| WeirdML | 49.9% | — |
| ALE-Bench | 944.17 | — |
Agentic & Tool Use Not comparable
Grok 4.3: 27.6 (#93), Llama2 70b Steerlm Chat: —
| Benchmark | Grok 4.3 | Llama2 70b Steerlm Chat |
|---|---|---|
| GDP.pdf | 8% | — |
| LMArena Search | 1165 | — |
| Vending-Bench 2 | 35.26 | — |
Reasoning Grok 4.3 leads
Grok 4.3: 34.9 (#60), Llama2 70b Steerlm Chat: 19.1 (#245)
| Benchmark | Grok 4.3 | Llama2 70b Steerlm Chat |
|---|---|---|
| LMArena Hard Prompts | 1397 | 1047 |
| 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 Grok 4.3 leads
Grok 4.3: 45.8 (#60), Llama2 70b Steerlm Chat: 28.8 (#223)
| Benchmark | Grok 4.3 | Llama2 70b Steerlm Chat |
|---|---|---|
| LMArena Math | 1388 | 1072 |
| FrontierMath (Tiers 1-3) | 42.8% | — |
| FrontierMath Tier 4 | 14.6% | — |
| OTIS Mock AIME 2024-2025 | 93.3% | — |
| ProofBench | 11% | — |
Knowledge Not comparable
Grok 4.3: 52.5 (#57), Llama2 70b Steerlm Chat: —
| Benchmark | Grok 4.3 | Llama2 70b Steerlm Chat |
|---|---|---|
| GPQA Diamond | 88.8% | — |
| SimpleQA Verified | 33.2% | — |
| LMArena Expert | 1385 | — |
Multimodal Not comparable
Grok 4.3: 31.6 (#96), Llama2 70b Steerlm Chat: —
| Benchmark | Grok 4.3 | Llama2 70b Steerlm Chat |
|---|---|---|
| LMArena Vision | 1231 | — |
| Blueprint-Bench 2 | 0% | — |
Multilingual Grok 4.3 leads
Grok 4.3: 50.7 (#114), Llama2 70b Steerlm Chat: 28.9 (#265)
| Benchmark | Grok 4.3 | Llama2 70b Steerlm Chat |
|---|---|---|
| LMArena Non-English | 1385 | 1064 |
| LMArena Chinese | 1423 | — |
| LMArena French | 1412 | — |
| LMArena German | 1395 | — |
| LMArena Japanese | 1379 | — |
| LMArena Korean | 1357 | — |
| LMArena Russian | 1399 | — |
| LMArena Spanish | 1402 | — |
Instruction Following Grok 4.3 leads
Grok 4.3: 72.3 (#135), Llama2 70b Steerlm Chat: 54.5 (#274)
| Benchmark | Grok 4.3 | Llama2 70b Steerlm Chat |
|---|---|---|
| LMArena Instruction Following | 1366 | 1061 |
Long Context Grok 4.3 leads
Grok 4.3: 42.5 (#118), Llama2 70b Steerlm Chat: 30.4 (#283)
| Benchmark | Grok 4.3 | Llama2 70b Steerlm Chat |
|---|---|---|
| LMArena Longer Query | 1393 | 998 |
Writing & Preference Grok 4.3 leads
Grok 4.3: 58.6 (#112), Llama2 70b Steerlm Chat: 31.8 (#278)
| Benchmark | Grok 4.3 | Llama2 70b Steerlm Chat |
|---|---|---|
| LMArena Text | 1397 | 1098 |
| LMArena Creative Writing | 1380 | 1091 |
| LMArena Multi-Turn | 1406 | 1058 |
| EQ-Bench 4 | 1075 | — |
Frequently asked questions
Is Grok 4.3 better than Llama2 70b Steerlm Chat?
Grok 4.3 is the stronger model overall, scoring 43.6 to 31.2 on the Noometry Index.
Is Grok 4.3 or Llama2 70b Steerlm Chat better for coding?
Grok 4.3 scores higher on coding benchmarks: 41.3 versus 29.3 in the Noometry coding category.
How many benchmarks do Grok 4.3 and Llama2 70b Steerlm Chat share?
9 benchmarks have published results for both models. Grok 4.3 has 40 scored results on Noometry and Llama2 70b Steerlm Chat has 9.