Workstation · 48 GB
70B Q4 serving begins here, with limited room for cache and concurrency.
Nex-N2.5-mini
The highest listed result may be a tight or offloaded fit. For daily use, prefer the first result marked Confortável and keep context modest.
vLLM
Linux GPU servers, concurrency and OpenAI-compatible production APIs. CUDA, ROCm and XPU feature coverage differs. Driver, shared-memory and quantization compatibility are version-specific.
Abrir guiaModelos open-weight representativos
Nex-N2.5-mini
English- Memória estimada
- ~21 GB
- Contexto
- 262,144 (config max_position_embeddings)
Qwen3.6 35B-A3B
- Memória estimada
- ~25 GB
- Contexto
- 262K native
Qwen3 32B
- Memória estimada
- ~24.5 GB
- Contexto
- 32K+
Qwen3 30B-A3B
- Memória estimada
- ~22.5 GB
- Contexto
- 32K+
Qwen3.6 27B
- Memória estimada
- ~18.5 GB
- Contexto
- 262K native
Qwen3.8 27B
- Memória estimada
- ~19.5 GB
- Contexto
- 262K native
Fit is based on estimated total model memory. Driver support, KV cache, multimodal projectors, concurrency, and desktop applications can all reduce available headroom. Current fit: Nex-N2.5-mini (Confortável), Qwen3.6 35B-A3B (Confortável), Qwen3 32B (Confortável), Qwen3 30B-A3B (Confortável), Qwen3.6 27B (Confortável), Qwen3.8 27B (Confortável).