Model comparison
Llama 3-8B vs Qwen3 235B-A22B
Qwen3 235B-A22B is the stronger model overall, scoring 43.2 to 25.3 on the Noometry Index. Llama 3-8B costs 8.7× less per token, which makes it the better buy when Qwen3 235B-A22B's lead doesn't matter for your workload.
Last verified . 24 shared benchmarks.
Meta
Alibaba (Qwen)Summary
- They share 24 benchmarks with published results for both. Llama 3-8B scores higher in 0 categories and Qwen3 235B-A22B in 8 categories; 8 gaps are clear of the uncertainty.
- The widest gap is in knowledge, where Qwen3 235B-A22B leads 49.5 to 7.8.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 1.9% for Llama 3-8B and 86.7% for Qwen3 235B-A22B.
- Llama 3-8B is cheaper at $0.14 / $0.14 per million input/output tokens, against $0.70 / $2.80 for Qwen3 235B-A22B.
- Qwen3 235B-A22B accepts more context: 131K tokens versus 8K.
Side by side
| Llama 3-8B | Qwen3 235B-A22B | |
|---|---|---|
| Provider | Meta | Alibaba (Qwen) |
| Noometry Index | 25.3 | 43.2 |
| Released | 2024-04-18 | 2025-04 |
| Weights | Open | Open |
| Context window | 8K | 131K |
| Max output | 8K | 16K |
| Input $ / M tokens | $0.14 | $0.70 |
| Output $ / M tokens | $0.14 | $2.80 |
| Results tracked | 34 | 49 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3 235B-A22B leads
Llama 3-8B: 30.5 (#285), Qwen3 235B-A22B: 43.8 (#72)
| Benchmark | Llama 3-8B | Qwen3 235B-A22B |
|---|---|---|
| LMArena Coding | 1152 | 1445 |
| Aider Polyglot | — | 59.6% |
| SciCode | — | 42.4% |
| WeirdML | — | 41% |
| BigCodeBench Instruct | 31.9% | — |
| BigCodeBench Complete | 36.9% | — |
| HumanEval+ | 56.7% | — |
| MBPP+ | 54.8% | — |
Agentic & Tool Use Not comparable
Llama 3-8B: —, Qwen3 235B-A22B: 33.7 (#48)
| Benchmark | Llama 3-8B | Qwen3 235B-A22B |
|---|---|---|
| Berkeley Function Calling Leaderboard | — | 52.1% |
| Vending-Bench 2 | — | -11.34 |
Reasoning Qwen3 235B-A22B leads
Llama 3-8B: 13.8 (#321), Qwen3 235B-A22B: 15.8 (#302)
| Benchmark | Llama 3-8B | Qwen3 235B-A22B |
|---|---|---|
| Chess Puzzles | 0% | 12% |
| LMArena Hard Prompts | 1133 | 1433 |
| DTBench | 43.9% | 80.3% |
| Epoch Capabilities Index | 116.33 | 143.87 |
| ForecastBench | 58.6 | 59.7 |
| ARC-AGI-2 | — | 1.3% |
| SimpleBench | — | 31% |
| Kagi LLM Benchmark | — | 69.4% |
| ARC-AGI-1 | — | 11% |
| CritPt | — | 0% |
| Mystery Game Puzzles | — | 9% |
| LMCA | — | 29.3% |
| Adversarial NLI | 57.3% | — |
| WinoGrande | 75.7% | — |
Math Qwen3 235B-A22B leads
Llama 3-8B: 8.3 (#317), Qwen3 235B-A22B: 48.5 (#51)
| Benchmark | Llama 3-8B | Qwen3 235B-A22B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 1.9% | 86.7% |
| LMArena Math | 1151 | 1431 |
| MATH Level 5 | 6.1% | 68.9% |
| Omni-MATH | — | 71.8% |
| FrontierMath (Feb 2025 set) | — | 8.5% |
| FrontierMath Tier 4 (v1) | — | 0% |
Knowledge Qwen3 235B-A22B leads
Llama 3-8B: 7.8 (#303), Qwen3 235B-A22B: 49.5 (#67)
| Benchmark | Llama 3-8B | Qwen3 235B-A22B |
|---|---|---|
| GPQA Diamond | 26.1% | 80.1% |
| LMArena Expert | 1113 | 1460 |
| SimpleQA Verified | — | 40.4% |
| MMLU-Pro | — | 84.4% |
| Confabulations | — | 15.6% |
| Vectara Hallucination Rate | — | 9.3% |
| GPQA (HELM) | — | 72.7% |
| ARC (AI2) Challenge | 82.8% | — |
| MMLU | 68.8% | — |
| OpenBookQA | 82.6% | — |
| TriviaQA | 67.7% | — |
Multilingual Qwen3 235B-A22B leads
Llama 3-8B: 31.0 (#256), Qwen3 235B-A22B: 52.5 (#84)
| Benchmark | Llama 3-8B | Qwen3 235B-A22B |
|---|---|---|
| LMArena Non-English | 1098 | 1410 |
| LMArena Chinese | 1076 | 1481 |
| LMArena French | 1158 | 1445 |
| LMArena German | 1104 | 1432 |
| LMArena Japanese | 967 | 1400 |
| LMArena Korean | 1003 | 1391 |
| LMArena Russian | 1109 | 1411 |
| LMArena Spanish | 1172 | 1429 |
Instruction Following Qwen3 235B-A22B leads
Llama 3-8B: 58.7 (#255), Qwen3 235B-A22B: 72.8 (#132)
| Benchmark | Llama 3-8B | Qwen3 235B-A22B |
|---|---|---|
| LMArena Instruction Following | 1127 | 1408 |
| IFEval | — | 83.5% |
Long Context Qwen3 235B-A22B leads
Llama 3-8B: 34.2 (#246), Qwen3 235B-A22B: 46.1 (#25)
| Benchmark | Llama 3-8B | Qwen3 235B-A22B |
|---|---|---|
| LMArena Longer Query | 1128 | 1426 |
| Fiction.LiveBench | — | 75% |
Writing & Preference Qwen3 235B-A22B leads
Llama 3-8B: 37.7 (#251), Qwen3 235B-A22B: 59.8 (#103)
| Benchmark | Llama 3-8B | Qwen3 235B-A22B |
|---|---|---|
| LMArena Text | 1166 | 1419 |
| LMArena Creative Writing | 1150 | 1384 |
| LMArena Multi-Turn | 1152 | 1432 |
| Short-Story Creative Writing | — | 83% |
| EQ-Bench Creative Writing | — | 1366 |
| WildBench | — | 86.6% |
Frequently asked questions
Is Llama 3-8B better than Qwen3 235B-A22B?
Qwen3 235B-A22B is the stronger model overall, scoring 43.2 to 25.3 on the Noometry Index. Llama 3-8B costs 8.7× less per token, which makes it the better buy when Qwen3 235B-A22B's lead doesn't matter for your workload.
Which is cheaper, Llama 3-8B or Qwen3 235B-A22B?
Llama 3-8B is cheaper. It lists at $0.14 per million input tokens and $0.14 per million output tokens; Qwen3 235B-A22B lists at $0.70 and $2.80.
Is Llama 3-8B or Qwen3 235B-A22B better for coding?
Qwen3 235B-A22B scores higher on coding benchmarks: 43.8 versus 30.5 in the Noometry coding category.
Which has the bigger context window?
Qwen3 235B-A22B does, with 131K tokens against 8K.
How many benchmarks do Llama 3-8B and Qwen3 235B-A22B share?
24 benchmarks have published results for both models. Llama 3-8B has 34 scored results on Noometry and Qwen3 235B-A22B has 49.