HARDWARE / VRAM
NVIDIA · 24 GB
The consumer sweet spot for 27B and 30B-A3B INT4/Q4 models.
RTX 3090 / 4090 class24 GB
01 / PRACTICAL CEILING
Nex-N2.5-mini
The highest listed result may be a tight or offloaded fit. For daily use, prefer the first result marked Confortable and keep context modest.
02 / DEFAULT RUNTIME
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.
Ouvrir le guide03 / MODEL ENVELOPE
Modèles open-weight représentatifs
Nex AGI
Nex-N2.5-mini
English35BQ4 estimated budget
- Mémoire estimée
- ~21 GB
- Contexte
- 262,144 (config max_position_embeddings)
Alibaba Qwen
Qwen3 30B-A3B
30B3B active · GGUF Q4 MoE
- Mémoire estimée
- ~22.5 GB
- Contexte
- 32K+
Alibaba Qwen
Qwen3.6 27B
27BQ4 / NVFP4 estimate
- Mémoire estimée
- ~18.5 GB
- Contexte
- 262K native
Alibaba Qwen
Qwen3.8 27B
27BQ4 estimate
- Mémoire estimée
- ~19.5 GB
- Contexte
- 262K native
Google
Gemma 3 27B
27BINT4
- Mémoire estimée
- ~20.5 GB
- Contexte
- 128K
Mistral AI
Devstral Small 2 24B
24BGGUF Q4 / BF16
- Mémoire estimée
- ~18.5 GB
- Contexte
- 256K
Attention
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 (Serré), Qwen3 30B-A3B (Serré), Qwen3.6 27B (Serré), Qwen3.8 27B (Serré), Gemma 3 27B (Serré), Devstral Small 2 24B (Serré).