Model comparison
Llama 3.1-70B vs Qwen Max
Qwen Max is the stronger model overall, scoring 34.7 to 29.6 on the Noometry Index. Llama 3.1-70B costs 7.0× less per token, which makes it the better buy when Qwen Max's lead doesn't matter for your workload.
Last verified . 20 shared benchmarks.
Meta
Alibaba (Qwen)Summary
- They share 20 benchmarks with published results for both. Llama 3.1-70B scores higher in 0 categories and Qwen Max in 8 categories; 7 gaps are clear of the uncertainty.
- The widest gap is in writing & preference, where Qwen Max leads 47.8 to 35.4.
- The biggest single-benchmark swing is MATH Level 5: 36.7% for Llama 3.1-70B and 67.2% for Qwen Max.
- Llama 3.1-70B is cheaper at $0.40 / $0.40 per million input/output tokens, against $1.60 / $6.40 for Qwen Max.
- Llama 3.1-70B accepts more context: 128K tokens versus 33K.
- Llama 3.1-70B has downloadable open weights; the other is API-only.
Side by side
| Llama 3.1-70B | Qwen Max | |
|---|---|---|
| Provider | Meta | Alibaba (Qwen) |
| Noometry Index | 29.6 | 34.7 |
| Released | 2024-07-23 | 2024-04-03 |
| Weights | Open | Proprietary |
| Context window | 128K | 33K |
| Max output | 4K | 8K |
| Input $ / M tokens | $0.40 | $1.60 |
| Output $ / M tokens | $0.40 | $6.40 |
| Results tracked | 35 | 23 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Too close to call
Llama 3.1-70B: 30.3 (#296), Qwen Max: 30.7 (#292)
| Benchmark | Llama 3.1-70B | Qwen Max |
|---|---|---|
| LMArena Coding | 1260 | 1288 |
| Aider Polyglot | — | 21.8% |
| WeirdML | 9% | — |
| BigCodeBench Instruct | 46.1% | — |
| BigCodeBench Complete | 54.8% | — |
Agentic & Tool Use Not comparable
Llama 3.1-70B: 25.1 (#112), Qwen Max: —
| Benchmark | Llama 3.1-70B | Qwen Max |
|---|---|---|
| TheAgentCompany | 6.9% | — |
| BALROG | 27.9% | — |
Reasoning Qwen Max leads
Llama 3.1-70B: 21.6 (#220), Qwen Max: 25.1 (#151)
| Benchmark | Llama 3.1-70B | Qwen Max |
|---|---|---|
| LMArena Hard Prompts | 1241 | 1269 |
| DTBench | 60% | — |
| LMCA | 14.8% | — |
| Epoch Capabilities Index | 125.92 | — |
Math Qwen Max leads
Llama 3.1-70B: 13.5 (#304), Qwen Max: 22.3 (#276)
| Benchmark | Llama 3.1-70B | Qwen Max |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 3.6% | 16.1% |
| LMArena Math | 1252 | 1275 |
| MATH Level 5 | 36.7% | 67.2% |
| Omni-MATH | 21% | — |
| FrontierMath (Feb 2025 set) | — | 1% |
Knowledge Qwen Max leads
Llama 3.1-70B: 24.2 (#269), Qwen Max: 30.3 (#228)
| Benchmark | Llama 3.1-70B | Qwen Max |
|---|---|---|
| GPQA Diamond | 44.2% | 56.1% |
| LMArena Expert | 1209 | 1248 |
| MMLU-Pro | 65.3% | — |
| GPQA (HELM) | 42.6% | — |
| MMLU | 80.1% | — |
Multilingual Qwen Max leads
Llama 3.1-70B: 38.8 (#225), Qwen Max: 41.8 (#202)
| Benchmark | Llama 3.1-70B | Qwen Max |
|---|---|---|
| LMArena Non-English | 1219 | 1263 |
| LMArena Chinese | 1215 | 1254 |
| LMArena French | 1261 | 1330 |
| LMArena German | 1222 | 1254 |
| LMArena Japanese | 1132 | 1205 |
| LMArena Korean | 1140 | 1142 |
| LMArena Russian | 1234 | 1274 |
| LMArena Spanish | 1253 | 1290 |
Instruction Following Qwen Max leads
Llama 3.1-70B: 65.3 (#223), Qwen Max: 66.5 (#208)
| Benchmark | Llama 3.1-70B | Qwen Max |
|---|---|---|
| LMArena Instruction Following | 1231 | 1262 |
| IFEval | 82.1% | — |
Long Context Qwen Max leads
Llama 3.1-70B: 37.6 (#214), Qwen Max: 39.4 (#180)
| Benchmark | Llama 3.1-70B | Qwen Max |
|---|---|---|
| LMArena Longer Query | 1241 | 1288 |
| Fiction.LiveBench | — | 66.7% |
Writing & Preference Qwen Max leads
Llama 3.1-70B: 35.4 (#267), Qwen Max: 47.8 (#205)
| Benchmark | Llama 3.1-70B | Qwen Max |
|---|---|---|
| LMArena Text | 1261 | 1282 |
| LMArena Creative Writing | 1232 | 1248 |
| LMArena Multi-Turn | 1256 | 1277 |
| EQ-Bench Creative Writing | 784 | — |
| WildBench | 75.8% | — |
Frequently asked questions
Is Llama 3.1-70B better than Qwen Max?
Qwen Max is the stronger model overall, scoring 34.7 to 29.6 on the Noometry Index. Llama 3.1-70B costs 7.0× less per token, which makes it the better buy when Qwen Max's lead doesn't matter for your workload.
Which is cheaper, Llama 3.1-70B or Qwen Max?
Llama 3.1-70B is cheaper. It lists at $0.40 per million input tokens and $0.40 per million output tokens; Qwen Max lists at $1.60 and $6.40.
Is Llama 3.1-70B or Qwen Max better for coding?
They score almost the same on coding (30.3 vs 30.7); test both on your own repository before choosing.
Which has the bigger context window?
Llama 3.1-70B does, with 128K tokens against 33K.
How many benchmarks do Llama 3.1-70B and Qwen Max share?
20 benchmarks have published results for both models. Llama 3.1-70B has 35 scored results on Noometry and Qwen Max has 23.