Llama 4 Maverick 17B Instruct (128E) is a high-capacity multimodal language model from Meta, built on a mixture-of-experts (MoE) architecture with 128 experts and 17 billion active parameters per forward...
| 信号 | 强度 | 权重 | 影响 |
|---|---|---|---|
| Benchmarksjust now | 70 | 30% | +21.0 |
| Pricingjust now | 99 | 15% | +14.9 |
| Capabilitiesjust now | 67 | 20% | +13.3 |
| Context Windowjust now | 86 | 10% | +8.6 |
| Recencyjust now | 52 | 15% | +7.8 |
| Output Capacityjust now | 70 | 10% | +7.0 |
LMSYS Arena Elo
1325
Percentile
70.8
Weight
30%
把当前模型放回同一服务商最近的发布节奏中查看。
Llama Guard 4 12B
coding
Llama 4 Maverick当前模型
coding
Llama 4 Scout
coding
Llama 3.3 70B Instruct (free)
coding
Llama 3.3 70B Instruct
coding
Llama 3.2 3B Instruct (free)
coding
Llama 3.2 3B Instruct
coding
Llama 3.2 1B Instruct
coding
社区和从业者反馈在基准测试和价格之上增加了真实世界的信号。
Share your experience with Llama 4 Maverick and help the community make better decisions.
LMSYS Arena Elo
1325
Percentile
70.8
Weight
30%
成本估算器
每月比类别平均节省$48.39