Model comparison
Claude 3 Haiku vs Grok 4.20 (Non-Reasoning)
Grok 4.20 (Non-Reasoning) is the stronger model overall, scoring 48.5 to 25.7 on the Noometry Index.
Last verified . 27 shared benchmarks.
Summary
- They share 27 benchmarks with published results for both. Claude 3 Haiku scores higher in 0 categories and Grok 4.20 (Non-Reasoning) in 9 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Grok 4.20 (Non-Reasoning) leads 48.0 to 9.2.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 1.8% for Claude 3 Haiku and 92.2% for Grok 4.20 (Non-Reasoning).
Side by side
| Claude 3 Haiku | Grok 4.20 (Non-Reasoning) | |
|---|---|---|
| Provider | Anthropic | xAI |
| Noometry Index | 25.7 | 48.5 |
| Released | 2024-03-07 | 2026-02-17 |
| Weights | Proprietary | Proprietary |
| Context window | — | 1M |
| Max output | — | 30K |
| Input $ / M tokens | — | $1.25 |
| Output $ / M tokens | — | $2.50 |
| Results tracked | 37 | 46 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Grok 4.20 (Non-Reasoning) leads
Claude 3 Haiku: 26.1 (#320), Grok 4.20 (Non-Reasoning): 41.8 (#104)
| Benchmark | Claude 3 Haiku | Grok 4.20 (Non-Reasoning) |
|---|---|---|
| WeirdML | 9.8% | 52.3% |
| LMArena Coding | 1199 | 1458 |
| LMArena WebDev | — | 1374 |
| BigCodeBench Instruct | 39.4% | — |
| BigCodeBench Complete | 50.1% | — |
| CadEval | 12% | — |
| ALE-Bench | — | 1,150 |
| HumanEval+ | 68.9% | — |
| MBPP+ | 68.8% | — |
Agentic & Tool Use Not comparable
Claude 3 Haiku: —, Grok 4.20 (Non-Reasoning): 34.3 (#43)
| Benchmark | Claude 3 Haiku | Grok 4.20 (Non-Reasoning) |
|---|---|---|
| Terminal-Bench | — | 57.3% |
| τ²-bench Banking | — | 18% |
| LMArena Search | — | 1189 |
| Vending-Bench 2 | — | 4,663 |
Reasoning Grok 4.20 (Non-Reasoning) leads
Claude 3 Haiku: 15.7 (#303), Grok 4.20 (Non-Reasoning): 52.2 (#30)
| Benchmark | Claude 3 Haiku | Grok 4.20 (Non-Reasoning) |
|---|---|---|
| Kagi LLM Benchmark | 34.2% | 75% |
| LMArena Hard Prompts | 1174 | 1451 |
| DTBench | 50.1% | 90.1% |
| LMCA | 8.8% | 38.7% |
| Epoch Capabilities Index | 118.3 | 152.01 |
| ForecastBench | 53.2 | 61.4 |
| ARC-AGI-2 | — | 65.1% |
| NYT Connections (extended) | — | 85.4% |
| ARC-AGI-1 | — | 89.5% |
| Chess Puzzles | — | 24% |
| Thematic Generalization | — | 63.8% |
| WinoGrande | 74.2% | — |
Math Grok 4.20 (Non-Reasoning) leads
Claude 3 Haiku: 9.2 (#313), Grok 4.20 (Non-Reasoning): 48.0 (#54)
| Benchmark | Claude 3 Haiku | Grok 4.20 (Non-Reasoning) |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.8% | 92.2% |
| LMArena Math | 1188 | 1455 |
| FrontierMath (Tiers 1-3) | — | 44.9% |
| FrontierMath Tier 4 | — | 17.1% |
| ProofBench | — | 14% |
| MATH Level 5 | 14.9% | — |
Knowledge Grok 4.20 (Non-Reasoning) leads
Claude 3 Haiku: 17.3 (#280), Grok 4.20 (Non-Reasoning): 52.8 (#54)
| Benchmark | Claude 3 Haiku | Grok 4.20 (Non-Reasoning) |
|---|---|---|
| GPQA Diamond | 36.3% | 89.3% |
| LMArena Expert | 1148 | 1440 |
| SimpleQA Verified | — | 30.2% |
| Confabulations | 34.2% | — |
| MMLU | 73.8% | — |
Multimodal Grok 4.20 (Non-Reasoning) leads
Claude 3 Haiku: 23.1 (#123), Grok 4.20 (Non-Reasoning): 33.2 (#88)
| Benchmark | Claude 3 Haiku | Grok 4.20 (Non-Reasoning) |
|---|---|---|
| LMArena Vision | 950 | 1263 |
| Blueprint-Bench 2 | — | 0% |
| LMArena Document | — | 1416 |
| ScienceQA | 72% | — |
Multilingual Grok 4.20 (Non-Reasoning) leads
Claude 3 Haiku: 36.1 (#238), Grok 4.20 (Non-Reasoning): 54.7 (#38)
| Benchmark | Claude 3 Haiku | Grok 4.20 (Non-Reasoning) |
|---|---|---|
| LMArena Non-English | 1178 | 1440 |
| LMArena Chinese | 1155 | 1481 |
| LMArena French | 1194 | 1477 |
| LMArena German | 1174 | 1463 |
| LMArena Japanese | 1102 | 1450 |
| LMArena Korean | 1108 | 1417 |
| LMArena Russian | 1204 | 1457 |
| LMArena Spanish | 1165 | 1442 |
Instruction Following Grok 4.20 (Non-Reasoning) leads
Claude 3 Haiku: 61.5 (#242), Grok 4.20 (Non-Reasoning): 75.0 (#79)
| Benchmark | Claude 3 Haiku | Grok 4.20 (Non-Reasoning) |
|---|---|---|
| LMArena Instruction Following | 1174 | 1420 |
Long Context Grok 4.20 (Non-Reasoning) leads
Claude 3 Haiku: 36.1 (#232), Grok 4.20 (Non-Reasoning): 45.4 (#33)
| Benchmark | Claude 3 Haiku | Grok 4.20 (Non-Reasoning) |
|---|---|---|
| LMArena Longer Query | 1191 | 1436 |
| CL-bench | — | 22.2% |
| CL-bench Life | — | 11.9% |
Writing & Preference Grok 4.20 (Non-Reasoning) leads
Claude 3 Haiku: 29.8 (#286), Grok 4.20 (Non-Reasoning): 65.9 (#42)
| Benchmark | Claude 3 Haiku | Grok 4.20 (Non-Reasoning) |
|---|---|---|
| LMArena Text | 1195 | 1451 |
| LMArena Creative Writing | 1157 | 1439 |
| EQ-Bench Creative Writing | 717 | 1574 |
| LMArena Multi-Turn | 1190 | 1456 |
Frequently asked questions
Is Claude 3 Haiku better than Grok 4.20 (Non-Reasoning)?
Grok 4.20 (Non-Reasoning) is the stronger model overall, scoring 48.5 to 25.7 on the Noometry Index.
Is Claude 3 Haiku or Grok 4.20 (Non-Reasoning) better for coding?
Grok 4.20 (Non-Reasoning) scores higher on coding benchmarks: 41.8 versus 26.1 in the Noometry coding category.
How many benchmarks do Claude 3 Haiku and Grok 4.20 (Non-Reasoning) share?
27 benchmarks have published results for both models. Claude 3 Haiku has 37 scored results on Noometry and Grok 4.20 (Non-Reasoning) has 46.