FITMYLLM · AUGUST 17, 2026 · OPEN METHODOLOGY · COMMUNITY BENCHMARKS
LIVE · UPDATED DAILY
Quantization Options Select your GPU for speed estimates Quant Bits VRAM @ 8K Quality IQ2_XXS 2.38 55.3 GB
54.0 + 1.3 KV
low IQ2_M 2.93 67.7 GB
66.4 + 1.3 KV
low Q2_K 3.16 72.8 GB
71.6 + 1.3 KV
low IQ3_XXS 3.25 74.9 GB
73.6 + 1.3 KV
low IQ3_XS 3.5 80.5 GB
79.2 + 1.3 KV
low Q3_K_S 3.64 83.6 GB
82.4 + 1.3 KV
low IQ3_M 3.76 86.3 GB
85.1 + 1.3 KV
low Q3_K_M 4 91.7 GB
90.5 + 1.3 KV
low Q3_K_L 4.3 98.5 GB
97.2 + 1.3 KV
moderate IQ4_XS 4.46 102.1 GB
100.8 + 1.3 KV
moderate Q4_K_S 4.67 106.8 GB
105.6 + 1.3 KV
moderate Q4_K_M 4.89 111.8 GB
110.5 + 1.3 KV
good Q5_K_S 5.57 127.1 GB
125.8 + 1.3 KV
good Q5_K_M 5.7 130.0 GB
128.7 + 1.3 KV
good Q6_K 6.56 149.3 GB
148.1 + 1.3 KV
excellent Q8_0 8.5 193.0 GB
191.7 + 1.3 KV
lossless FP16 16 361.7 GB
360.5 + 1.3 KV
lossless
Select your GPU above to see speed estimates and compatibility for each quantization.
Too big for a single GPU — plan a multi-GPU deployment
Even the lightest quant needs ~55 GB. Size GPUs, replicas, TCO and scaling for a production setup. Open in Enterprise →
▸ READY TO RUN THIS? RENT BY THE HOUR
RENT A GPU AND RUN FALCON 180B NOW
Spin up an A100 / H100 / 4090 in ~60s. Pay by the second. Cancel anytime.
Community Ratings Chat Coding Reasoning Creative Vision Roleplay Agentic
Loading ratings...
Run this model IQ2_XXS — 54.0 GB VRAM IQ2_M — 66.4 GB VRAM Q2_K — 71.6 GB VRAM IQ3_XXS — 73.6 GB VRAM IQ3_XS — 79.2 GB VRAM Q3_K_S — 82.4 GB VRAM IQ3_M — 85.1 GB VRAM Q3_K_M — 90.5 GB VRAM Q3_K_L — 97.2 GB VRAM IQ4_XS — 100.8 GB VRAM Q4_K_S — 105.6 GB VRAM Q4_K_M — 110.5 GB VRAM Q5_K_S — 125.8 GB VRAM Q5_K_M — 128.7 GB VRAM Q6_K — 148.1 GB VRAM Q8_0 — 191.7 GB VRAM FP16 — 360.5 GB VRAM
Ollama llama.cpp vLLM LM Studio KoboldCpp Jan Docker
▸ Easiest way to get started · Beginners
DOCS ↗ curl -fsSL https://ollama.com/install.sh | shCOPY
$ ollama run falcon:180bCOPY
Tag may need adjustment — check ollama.com/library/falcon for available tags.
▸ SETUP GUIDE >_
Auto-setup with fitmyllm CLI Detects your GPU, recommends the best model, downloads it, and starts chatting — zero config. Benchmarks your speed and contributes anonymous data to improve predictions.
Auto-detect GPU Live tok/s in chat Speed benchmarks 9 inference engines
GPUs that can run this model At Q4_K_M quantization. Sorted by minimum VRAM.
Find the best GPU for Falcon 180B
Build Hardware for Falcon 180B Falcon 180B — massive model from TII, requires multi-GPU setup.
Read full model card ▸ COLOPHON FITMYLLM · INDEPENDENT · DATA-DRIVEN
FITMYLLM · EST. 2025 · © 2026
RECOMMENDATIONS FROM PUBLISHED MATH, CORRECTED BY THE COMMUNITY — 30.
▸ SPEC SHEET
Falcon 180B — 180B Dense. ▸ SPECIFICATIONS
PARAMETERS 180B
ARCHITECTURE Dense Transformer
CONTEXT LENGTH 8K tokens
CAPABILITIES chat
RELEASE DATE 2023-09-06
PROVIDER TII
FAMILY falcon ▸ VRAM REQUIREMENTS
QUANT BPW VRAM QUALITY IQ2_XXS 2.38 54.0 GB 65% IQ2_M 2.93 66.4 GB 75% Q2_K 3.16 71.6 GB 78% IQ3_XXS 3.25 73.6 GB 82% IQ3_XS 3.5 79.2 GB 84% Q3_K_S 3.64 82.4 GB 85% IQ3_M 3.76 85.1 GB 86% Q3_K_M 4 90.5 GB 88% Q3_K_L 4.3 97.2 GB 90% IQ4_XS 4.46 100.8 GB 92% Q4_K_S 4.67 105.6 GB 93% Q4_K_M 4.89 110.5 GB 94% Q5_K_S 5.57 125.8 GB 96% Q5_K_M 5.7 128.7 GB 96% Q6_K 6.56 148.1 GB 97% Q8_0 8.5 191.7 GB 100% FP16 16 360.5 GB 100%
§ 01 BENCHMARK SCORES
HumanEval 54.0
MMLU-PRO 45.0
IFEval 41.0
BBH 65.0
Arena Elo 1049.0
§ 02 RUN COMMAND
Run Falcon 180B locally with Ollama — needs 110.5 GB VRAM at Q4_K_M:
$ ollama run falcon:180b
§ 03 COMPATIBLE GPUs
30 @ Q4_K_M Feedback 