Falcon-H1R-7B
Files and memory at 8K context
| Quant | File | Files | + KV + runtime | Smallest fit | Source |
|---|---|---|---|---|---|
| Q4_K_M tiiuae/Falcon-H1R-7B-GGUF | 4.28 GiB | 1 | 5.6 GiB+ | 8 GB GPU | HF files |
| Q5_K_M tiiuae/Falcon-H1R-7B-GGUF | 5.02 GiB | 1 | 6.4 GiB+ | 8 GB GPU | HF files |
| Q8_0 tiiuae/Falcon-H1R-7B-GGUF | 7.51 GiB | 1 | 8.9 GiB+ | 12 GB GPU | HF files |
| FP16 tiiuae/Falcon-H1R-7B-GGUF | 14.14 GiB | 1 | 15.5 GiB+ | 16 GB GPU | HF files |
| MLX 4-bit mlx-community/Falcon-H1R-7B-4bit | 3.98 GiB | 1 | 5.3 GiB+ | 8 GB Mac | HF files |
| MLX 8-bit mlx-community/Falcon-H1R-7B-8bit | 7.51 GiB | 2 | 8.9 GiB+ | 16 GB Mac | HF files |
Sizes are exact file bytes (text weights; projectors listed separately). “+” = recurrent state or other terms unknown. Smallest fit adds 0.5 GiB spare; Macs use 75% of memory.
Quality
Artificial Analysis Intelligence Index
7.8 AA estimate
Variant: Falcon-H1R-7B · observed 2026-10-17 · source
Rank #71 of 75 researched 2026 models.
Arena text rating
Unknown
Scores describe the evaluated hosted variant, not each quantized file. AA and Arena are never averaged.
Reported speeds 0
No reported measurement yet. Submit one.
Community threads
- Function call and tool usage with Open WebUI Hugging Face · 2026-01-19
- How to control reasoning effort in Falcon H1R 7B? Hugging Face · 2026-09-13
Qualitative signals only; titles are not verified claims.