Model comparison
Llama 3.1-70B vs Llama 4 Maverick
Llama 4 Maverick is the stronger model overall, scoring 30.6 to 29.4 on the Noometry Index.
Last verified . 32 shared benchmarks.
MetaSummary
- They share 32 benchmarks with published results for both. Llama 3.1-70B scores higher in 3 categories and Llama 4 Maverick in 6 categories; 9 gaps are clear of the uncertainty.
- The widest gap is in math, where Llama 4 Maverick leads 24.6 to 12.7.
- The biggest single-benchmark swing is MATH Level 5: 36.7% for Llama 3.1-70B and 73% for Llama 4 Maverick.
- Llama 4 Maverick is cheaper at $0.19 / $0.65 per million input/output tokens, against $0.40 / $0.40 for Llama 3.1-70B.
Side by side
| Llama 3.1-70B | Llama 4 Maverick | |
|---|---|---|
| Provider | Meta | Meta |
| Noometry Index | 29.4 | 30.6 |
| Released | 2024-07-23 | 2025-04-05 |
| Weights | Open | Open |
| Context window | 128K | 128K |
| Max output | 4K | 4K |
| Input $ / M tokens | $0.40 | $0.19 |
| Output $ / M tokens | $0.40 | $0.65 |
| Results tracked | 35 | 54 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Llama 3.1-70B leads
Llama 3.1-70B: 29.8 (#292), Llama 4 Maverick: 26.3 (#318)
| Benchmark | Llama 3.1-70B | Llama 4 Maverick |
|---|---|---|
| WeirdML | 9% | 24.5% |
| BigCodeBench Instruct | 46.1% | 49.7% |
| LMArena Coding | 1260 | 1302 |
| BigCodeBench Complete | 54.8% | 61.4% |
| SWE-bench Verified (bash only) | — | 21% |
| Aider Polyglot | — | 15.6% |
| SciCode | — | 33.1% |
| ALE-Bench | — | 172.97 |
Agentic & Tool Use Llama 4 Maverick leads
Llama 3.1-70B: 24.9 (#104), Llama 4 Maverick: 27.9 (#86)
| Benchmark | Llama 3.1-70B | Llama 4 Maverick |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 37.3% |
| TheAgentCompany | 6.9% | — |
| BALROG | 27.9% | — |
Reasoning Llama 3.1-70B leads
Llama 3.1-70B: 20.8 (#215), Llama 4 Maverick: 10.1 (#337)
| Benchmark | Llama 3.1-70B | Llama 4 Maverick |
|---|---|---|
| LMArena Hard Prompts | 1241 | 1281 |
| DTBench | 60% | 61.9% |
| LMCA | 14.8% | 15.9% |
| Epoch Capabilities Index | 125.91 | 132.2 |
| ARC-AGI-2 | — | 0% |
| SimpleBench | — | 27.7% |
| Kagi LLM Benchmark | — | 55.9% |
| NYT Connections (extended) | — | 8% |
| ARC-AGI-1 | — | 4.4% |
| CritPt | — | 0% |
| EnigmaEval | — | 0.6% |
| ForecastBench | — | 57.5 |
Math Llama 4 Maverick leads
Llama 3.1-70B: 12.7 (#298), Llama 4 Maverick: 24.6 (#256)
| Benchmark | Llama 3.1-70B | Llama 4 Maverick |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 3.6% | 20.6% |
| Omni-MATH | 21% | 42.2% |
| LMArena Math | 1252 | 1299 |
| MATH Level 5 | 36.7% | 73% |
| FrontierMath (Feb 2025 set) | — | 0.7% |
Knowledge Llama 4 Maverick leads
Llama 3.1-70B: 24.1 (#264), Llama 4 Maverick: 33.3 (#196)
| Benchmark | Llama 3.1-70B | Llama 4 Maverick |
|---|---|---|
| GPQA Diamond | 44.2% | 67% |
| MMLU-Pro | 65.3% | 81% |
| GPQA (HELM) | 42.6% | 65% |
| LMArena Expert | 1209 | 1259 |
| Humanity's Last Exam | — | 5.7% |
| Confabulations | — | 22.6% |
| Vectara Hallucination Rate | — | 8.2% |
| MMLU | 80.1% | — |
Multimodal Not comparable
Llama 3.1-70B: —, Llama 4 Maverick: 30.9 (#100)
| Benchmark | Llama 3.1-70B | Llama 4 Maverick |
|---|---|---|
| LMArena Vision | — | 1141 |
| GeoBench | — | 52% |
| SpatialViz-Bench | — | 31.8% |
Multilingual Llama 4 Maverick leads
Llama 3.1-70B: 38.9 (#220), Llama 4 Maverick: 42.4 (#190)
| Benchmark | Llama 3.1-70B | Llama 4 Maverick |
|---|---|---|
| LMArena Non-English | 1219 | 1269 |
| LMArena Chinese | 1215 | 1277 |
| LMArena French | 1260 | 1258 |
| LMArena German | 1222 | 1291 |
| LMArena Japanese | 1132 | 1207 |
| LMArena Korean | 1140 | 1203 |
| LMArena Russian | 1234 | 1286 |
| LMArena Spanish | 1253 | 1293 |
Instruction Following Llama 4 Maverick leads
Llama 3.1-70B: 65.5 (#218), Llama 4 Maverick: 71.9 (#140)
| Benchmark | Llama 3.1-70B | Llama 4 Maverick |
|---|---|---|
| IFEval | 82.1% | 90.8% |
| LMArena Instruction Following | 1231 | 1267 |
Long Context Llama 3.1-70B leads
Llama 3.1-70B: 37.6 (#208), Llama 4 Maverick: 31.4 (#274)
| Benchmark | Llama 3.1-70B | Llama 4 Maverick |
|---|---|---|
| LMArena Longer Query | 1241 | 1280 |
| Fiction.LiveBench | — | 46.2% |
Writing & Preference Llama 4 Maverick leads
Llama 3.1-70B: 35.6 (#262), Llama 4 Maverick: 39.0 (#247)
| Benchmark | Llama 3.1-70B | Llama 4 Maverick |
|---|---|---|
| LMArena Text | 1261 | 1287 |
| LMArena Creative Writing | 1232 | 1266 |
| EQ-Bench Creative Writing | 784 | 860 |
| WildBench | 75.8% | 80% |
| LMArena Multi-Turn | 1256 | 1289 |
| Short-Story Creative Writing | — | 62% |
Frequently asked questions
Is Llama 3.1-70B better than Llama 4 Maverick?
Llama 4 Maverick is the stronger model overall, scoring 30.6 to 29.4 on the Noometry Index.
Which is cheaper, Llama 3.1-70B or Llama 4 Maverick?
Llama 4 Maverick is cheaper. It lists at $0.19 per million input tokens and $0.65 per million output tokens; Llama 3.1-70B lists at $0.40 and $0.40.
Is Llama 3.1-70B or Llama 4 Maverick better for coding?
Llama 3.1-70B scores higher on coding benchmarks: 29.8 versus 26.3 in the Noometry coding category.
Which has the bigger context window?
Both accept 128K tokens.
How many benchmarks do Llama 3.1-70B and Llama 4 Maverick share?
32 benchmarks have published results for both models. Llama 3.1-70B has 35 scored results on Noometry and Llama 4 Maverick has 54.