HARDWARE / VRAM
Datacenter · 80 GB
Large MoE models and high-throughput servers, still bounded by quantization and cache.
H100 / H200 / MI300X class80 GB
01 / PRACTICAL CEILING
Ling-3.0-flash
The highest listed result may be a tight or offloaded fit. For daily use, prefer the first result marked Comfortable and keep context modest.
02 / DEFAULT RUNTIME
TensorRT-LLM
Maximum NVIDIA throughput after engine tuning. NVIDIA-only and sensitive to supported GPU, CUDA, model architecture and engine-build settings.
Open guide03 / MODEL ENVELOPE
Representative open-weight models
inclusionAI
Ling-3.0-flash
124BQ4 estimated budget
- Estimated memory
- ~74.4 GB
- Context
- Check selected runtime and checkpoint
Meta
Llama 3.1 70B
70BGGUF Q4
- Estimated memory
- ~46 GB
- Context
- 128K
Nex AGI
Nex-N2.5-mini
35BQ4 estimated budget
- Estimated memory
- ~21 GB
- Context
- 262,144 (config max_position_embeddings)
Alibaba Qwen
Qwen3.6 35B-A3B
35B3B active · Q4 / NVFP4 MoE
- Estimated memory
- ~25 GB
- Context
- 262K native
Alibaba Qwen
Qwen3 32B
32BGGUF Q4
- Estimated memory
- ~24.5 GB
- Context
- 32K+
Alibaba Qwen
Qwen3 30B-A3B
30B3B active · GGUF Q4 MoE
- Estimated memory
- ~22.5 GB
- Context
- 32K+
Watch out
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: Ling-3.0-flash (Tight fit), Llama 3.1 70B (Comfortable), Nex-N2.5-mini (Comfortable), Qwen3.6 35B-A3B (Comfortable), Qwen3 32B (Comfortable), Qwen3 30B-A3B (Comfortable).