Can Unlimited OCR 2 run on the A100 40GB?
Yes. At 8K context the best fit is F16, using about 6.6 GB of the 38.0 GB usable on the A100 40GB. Estimated speed ceiling: ~769 tokens/sec.
Memory by quant and context
| Quant | 4K ctx | 8K ctx | 32K ctx | 128K ctx | Max tok/s (est., 8K) |
|---|---|---|---|---|---|
| 2-bit (Q2_K) | 2.0 GB Fits | 2.3 GB Fits | 3.7 GB Fits | 9.3 GB Fits | 3609 |
| 3-bit (Q3_K_M) | 2.4 GB Fits | 2.6 GB Fits | 4.0 GB Fits | 9.7 GB Fits | 2769 |
| 4-bit (Q4_K_M) | 2.8 GB Fits | 3.0 GB Fits | 4.4 GB Fits | 10.1 GB Fits | 2214 |
| 5-bit (Q5_K_M) | 3.1 GB Fits | 3.3 GB Fits | 4.7 GB Fits | 10.4 GB Fits | 1900 |
| 6-bit (Q6_K) • | 3.4 GB Fits | 3.7 GB Fits | 5.1 GB Fits | 10.7 GB Fits | 1650 |
| 8-bit (Q8_0) • | 3.8 GB Fits | 4.0 GB Fits | 5.4 GB Fits | 11.1 GB Fits | 1444 |
| 16-bit (F16/BF16) • | 6.4 GB Fits | 6.6 GB Fits | 8.0 GB Fits | 13.6 GB Fits | 769 |
• real GGUF file size from AtomicChat/Unlimited-OCR-2-GGUF. Speed is a bandwidth ceiling (1555 GB/s), not a benchmark. Methodology.
What fits instead
- Cheapest listed GPU that runs Unlimited OCR 2: Arc B580 (12 GB, launch MSRP $249) at F16, fully in VRAM at 8K context.
- Largest popular model that fits the A100 40GB: Qwen3.6 35B A3B (36B, MoE) at Q6_K, about 29.0 GB at 8K context.
"Runs" and "fits" mean the whole model in GPU memory at 8K context (4-bit or better for the model pick). Prices are US launch MSRPs, not today's street prices. "Popular" means the 60 most-downloaded models we track.
More
Unlimited OCR 2 on every GPU and Mac · Every model that runs on the A100 40GB · Other models that fit 48 GB VRAM · Try other settings in the calculator
Guides: Best GPU for local LLMs by budget · Can I run gpt-oss, Llama 70B or Qwen3? · Mac vs PC for local LLMs
Data: Hugging Face Hub (baidu/Unlimited-OCR-2), NVIDIA specs. Model data last checked 2026-11-17; refreshed hourly.