Can Falcon H1R 7B run on the RTX 5080?
Yes. At 8K context the best fit is Q8_0, using about 8.5 GB of the 15.2 GB usable on the RTX 5080. Estimated speed ceiling: ~119 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) • | 3.5 GB Fits | 3.7 GB Fits | 4.7 GB Fits | 8.8 GB Fits | 332 |
| 3-bit (Q3_K_M) • | 4.3 GB Fits | 4.4 GB Fits | 5.5 GB Fits | 9.6 GB Fits | 260 |
| 4-bit (Q4_K_M) • | 5.1 GB Fits | 5.3 GB Fits | 6.3 GB Fits | 10.4 GB Fits | 209 |
| 5-bit (Q5_K_M) • | 5.8 GB Fits | 6.0 GB Fits | 7.0 GB Fits | 11.2 GB Fits | 178 |
| 6-bit (Q6_K) • | 6.6 GB Fits | 6.8 GB Fits | 7.8 GB Fits | 12.0 GB Fits | 154 |
| 8-bit (Q8_0) • | 8.3 GB Fits | 8.5 GB Fits | 9.5 GB Fits | 13.7 GB Fits | 119 |
| 16-bit (F16/BF16) • | 15.0 GB Tight | 15.1 GB Tight | 16.2 GB No | 20.3 GB No | 63 |
• real GGUF file size from unsloth/Falcon-H1R-7B-GGUF. Speed is a bandwidth ceiling (960 GB/s), not a benchmark. Methodology.
More
Falcon H1R 7B on every GPU and Mac · Every model that runs on the RTX 5080 · Other models that fit 16 GB VRAM · Try other settings in the calculator
Data: Hugging Face Hub (tiiuae/Falcon-H1R-7B), NVIDIA specs. Data as of 2026-11-17.