Can I run LongCat 2.0 locally?
LongCat 2.0 by meituan-longcat has 1.78T parameters (1.78T active per token, 768 experts, 12 used per token). At 4-bit with an 8K context it needs about 1012 GB of GPU memory. It fits on 0 of 65 GPUs and Macs we track.
Memory needed by quant and context length
GB of GPU memory = weights + KV cache (FP16) + runtime overhead. Weights use parameter count × bits per weight. How we calculate.
| Quant | Quality | 4K ctx | 8K ctx | 32K ctx | 128K ctx |
|---|---|---|---|---|---|
| 2-bit (Q2_K) | rough | 621 GB | 621 GB | 622 GB | 627 GB |
| 3-bit (Q3_K_M) | usable | 809 GB | 809 GB | 811 GB | 815 GB |
| 4-bit (Q4_K_M) | good, most popular | 1012 GB | 1012 GB | 1013 GB | 1018 GB |
| 5-bit (Q5_K_M) | very good | 1179 GB | 1179 GB | 1181 GB | 1185 GB |
| 6-bit (Q6_K) | near lossless | 1357 GB | 1357 GB | 1358 GB | 1363 GB |
| 8-bit (Q8_0) | effectively lossless | 1758 GB | 1758 GB | 1759 GB | 1764 GB |
| 16-bit (F16/BF16) | original weights | 3308 GB | 3308 GB | 3310 GB | 3314 GB |
Which GPUs and Macs run LongCat 2.0?
Best quality quant that fits at 8K context. "Tight" = within 10% of usable memory. Speed is a ceiling from memory bandwidth (real speed is typically 50–80% of it).
Prefer an API?
Meituan: LongCat 2.0 is on OpenRouter at $0.30 per 1M input tokens and $1.20 per 1M output tokens. Price details and history →
Questions
How much VRAM does LongCat 2.0 need?
About 1012 GB at 4-bit (Q4_K_M) with an 8K context: 1011 GB of weights (estimated), 0.4 GB of KV cache and 0.7 GB of runtime overhead. At 8-bit it needs about 1758 GB.
What is the smallest GPU that runs LongCat 2.0?
No single consumer GPU in our table fits it at Q4_K_M with 8K context.
Can I run LongCat 2.0 on a Mac?
Not at Q4_K_M with 8K context on any Mac in our table, even a 512 GB Mac Studio, using the default GPU memory limit.
Model facts
| Hugging Face repo | meituan-longcat/LongCat-2.0 |
| Parameters | 1,775,560,491,136 (1.78T) · ~1.78T active (computed from config) |
| Layers / hidden size | 38 / 8192 |
| Attention | MLA; 64 heads, 64 KV heads, head dim 128 |
| KV cache per token (FP16) | 52.3 KiB |
| Max context | 262,144 tokens |
| License | mit |
| Released on HF | Jul 5, 2026 |
| Downloads (30d) / likes | 3,058 / 286 |
Sources: Hugging Face model API · config.json (meituan-longcat/LongCat-2.0) · GPU specs from vendor pages (list). Data fetched Nov 17, 2026. Spot an error? Report it.