Qwen3 4B Instruct 2507
Código abiertopor Alibaba · Publicado el 2025-08-05
47.2
puntuación promedio
N/A
Precio de entrada
N/A
Precio de salida
N/A
Ventana de contexto
text-generation
Tipo
Tested on 6 benchmarks with 47.2% average. Top scores: OpenCompass — IFEval (82.4%), OpenCompass — MMLU-Pro (63.0%), OpenCompass — GPQA-Diamond (52.3%).
Puntuaciones de benchmark
| Benchmark | Categoría | Puntuación | Bar |
|---|---|---|---|
| OpenCompass — IFEval | language | 82.4 | |
| OpenCompass — MMLU-Pro | knowledge | 63.0 | |
| OpenCompass — GPQA-Diamond | knowledge | 52.3 | |
| OpenCompass — AIME2025 | math | 46.9 | |
| OpenCompass — LiveCodeBenchV6 | coding | 33.5 | |
| OpenCompass — HLE | knowledge | 5.1 |
Modelos similares
Qwen2 VL 7B InstructAlibaba
47.3
Gemini 1.5 Flash (May 2024)Google DeepMind
47.4
Llama 3.3 70B InstructMeta
46.9
gpt-oss-120bOpenAI
46.9