Model comparison
Llama 4 Maverick vs Qwen3 8B
Qwen3 8B is the stronger model overall, scoring 33.7 to 30.9 on the Noometry Index.
Last verified . 10 shared benchmarks.
Meta
Alibaba (Qwen)Summary
- They share 10 benchmarks with published results for both. Llama 4 Maverick scores higher in 0 categories and Qwen3 8B in 6 categories; 6 gaps are clear of the uncertainty.
- The widest gap is in math, where Qwen3 8B leads 34.9 to 26.0.
- The biggest single-benchmark swing is OTIS Mock AIME 2024-2025: 20.6% for Llama 4 Maverick and 56.1% for Qwen3 8B.
- Both cost about the same: $0.19 input and $0.65 output per million tokens.
- Qwen3 8B accepts more context: 131K tokens versus 128K.
Side by side
| Llama 4 Maverick | Qwen3 8B | |
|---|---|---|
| Provider | Meta | Alibaba (Qwen) |
| Noometry Index | 30.9 | 33.7 |
| Released | 2025-04-05 | 2025-04 |
| Weights | Open | Open |
| Context window | 128K | 131K |
| Max output | 4K | 8K |
| Input $ / M tokens | $0.19 | $0.18 |
| Output $ / M tokens | $0.65 | $0.70 |
| Results tracked | 54 | 11 |
Sponsored placements are available on pages like this one. Advertise on Noometry
Category by category
Coding Qwen3 8B leads
Llama 4 Maverick: 26.6 (#324), Qwen3 8B: 34.0 (#248)
| Benchmark | Llama 4 Maverick | Qwen3 8B |
|---|---|---|
| SciCode | 33.1% | 22.6% |
| SWE-bench Verified (bash only) | 21% | — |
| Aider Polyglot | 15.6% | — |
| WeirdML | 24.5% | — |
| BigCodeBench Instruct | 49.7% | — |
| LMArena Coding | 1302 | — |
| BigCodeBench Complete | 61.4% | — |
| ALE-Bench | 172.97 | — |
Agentic & Tool Use Qwen3 8B leads
Llama 4 Maverick: 28.2 (#91), Qwen3 8B: 30.2 (#78)
| Benchmark | Llama 4 Maverick | Qwen3 8B |
|---|---|---|
| Berkeley Function Calling Leaderboard | 37.3% | 42.6% |
Reasoning Qwen3 8B leads
Llama 4 Maverick: 10.1 (#342), Qwen3 8B: 16.6 (#303)
| Benchmark | Llama 4 Maverick | Qwen3 8B |
|---|---|---|
| CritPt | 0% | 0% |
| DTBench | 61.9% | 59.7% |
| LMCA | 15.9% | 8.8% |
| Epoch Capabilities Index | 132.2 | 136.17 |
| ARC-AGI-2 | 0% | — |
| SimpleBench | 27.7% | — |
| Kagi LLM Benchmark | 55.9% | — |
| NYT Connections (extended) | 8% | — |
| ARC-AGI-1 | 4.4% | — |
| Chess Puzzles | — | 5% |
| EnigmaEval | 0.6% | — |
| LMArena Hard Prompts | 1281 | — |
| ForecastBench | 57.5 | — |
Math Qwen3 8B leads
Llama 4 Maverick: 26.0 (#262), Qwen3 8B: 34.9 (#191)
| Benchmark | Llama 4 Maverick | Qwen3 8B |
|---|---|---|
| OTIS Mock AIME 2024-2025 | 20.6% | 56.1% |
| Omni-MATH | 42.2% | — |
| LMArena Math | 1299 | — |
| MATH Level 5 | 73% | — |
| FrontierMath (Feb 2025 set) | 0.7% | — |
Knowledge Qwen3 8B leads
Llama 4 Maverick: 33.4 (#204), Qwen3 8B: 36.1 (#173)
| Benchmark | Llama 4 Maverick | Qwen3 8B |
|---|---|---|
| GPQA Diamond | 67% | 56.8% |
| Vectara Hallucination Rate | 8.2% | 4.8% |
| Humanity's Last Exam | 5.7% | — |
| MMLU-Pro | 81% | — |
| Confabulations | 22.6% | — |
| GPQA (HELM) | 65% | — |
| LMArena Expert | 1259 | — |
Multimodal Not comparable
Llama 4 Maverick: 31.6 (#105), Qwen3 8B: —
| Benchmark | Llama 4 Maverick | Qwen3 8B |
|---|---|---|
| LMArena Vision | 1142 | — |
| GeoBench | 52% | — |
| SpatialViz-Bench | 31.8% | — |
Multilingual Not comparable
Llama 4 Maverick: 42.2 (#195), Qwen3 8B: —
| Benchmark | Llama 4 Maverick | Qwen3 8B |
|---|---|---|
| LMArena Non-English | 1269 | — |
| LMArena Chinese | 1277 | — |
| LMArena French | 1259 | — |
| LMArena German | 1291 | — |
| LMArena Japanese | 1207 | — |
| LMArena Korean | 1203 | — |
| LMArena Russian | 1286 | — |
| LMArena Spanish | 1293 | — |
Instruction Following Not comparable
Llama 4 Maverick: 71.7 (#146), Qwen3 8B: —
| Benchmark | Llama 4 Maverick | Qwen3 8B |
|---|---|---|
| IFEval | 90.8% | — |
| LMArena Instruction Following | 1267 | — |
Long Context Qwen3 8B leads
Llama 4 Maverick: 31.4 (#279), Qwen3 8B: 37.9 (#210)
| Benchmark | Llama 4 Maverick | Qwen3 8B |
|---|---|---|
| Fiction.LiveBench | 46.2% | 62.1% |
| LMArena Longer Query | 1280 | — |
Writing & Preference Not comparable
Llama 4 Maverick: 38.8 (#252), Qwen3 8B: —
| Benchmark | Llama 4 Maverick | Qwen3 8B |
|---|---|---|
| LMArena Text | 1287 | — |
| LMArena Creative Writing | 1267 | — |
| Short-Story Creative Writing | 62% | — |
| EQ-Bench Creative Writing | 860 | — |
| WildBench | 80% | — |
| LMArena Multi-Turn | 1289 | — |
Frequently asked questions
Is Llama 4 Maverick better than Qwen3 8B?
Qwen3 8B is the stronger model overall, scoring 33.7 to 30.9 on the Noometry Index.
Which is cheaper, Llama 4 Maverick or Qwen3 8B?
Llama 4 Maverick is cheaper. It lists at $0.19 per million input tokens and $0.65 per million output tokens; Qwen3 8B lists at $0.18 and $0.70.
Is Llama 4 Maverick or Qwen3 8B better for coding?
Qwen3 8B scores higher on coding benchmarks: 34.0 versus 26.6 in the Noometry coding category.
Which has the bigger context window?
Qwen3 8B does, with 131K tokens against 128K.
How many benchmarks do Llama 4 Maverick and Qwen3 8B share?
10 benchmarks have published results for both models. Llama 4 Maverick has 54 scored results on Noometry and Qwen3 8B has 11.