FITMYLLM · AUGUST 17, 2026 · OPEN METHODOLOGY · COMMUNITY BENCHMARKS
LIVE · UPDATED DAILY
▸ READY TO RUN THIS? RENT BY THE HOUR
RENT A GPU AND RUN NEMOTRON-4 340B 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 — 101.6 GB VRAM IQ2_M — 125.0 GB VRAM Q2_K — 134.8 GB VRAM IQ3_XXS — 138.6 GB VRAM IQ3_XS — 149.2 GB VRAM Q3_K_S — 155.2 GB VRAM IQ3_M — 160.3 GB VRAM Q3_K_M — 170.5 GB VRAM Q3_K_L — 183.2 GB VRAM IQ4_XS — 190.0 GB VRAM Q4_K_S — 199.0 GB VRAM Q4_K_M — 208.3 GB VRAM Q5_K_S — 237.2 GB VRAM Q5_K_M — 242.7 GB VRAM Q6_K — 279.3 GB VRAM Q8_0 — 361.7 GB VRAM FP16 — 680.5 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 nemotron:340b-q4_K_MCOPY
Tag may need adjustment — check ollama.com/library/nemotron 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 Nemotron-4 340B
Build Hardware for Nemotron-4 340B Nemotron-4 340B — NVIDIA's massive model for research and enterprise.
Read full model card ▸ COLOPHON FITMYLLM · INDEPENDENT · DATA-DRIVEN
FITMYLLM · EST. 2025 · © 2026
RECOMMENDATIONS FROM PUBLISHED MATH, CORRECTED BY THE COMMUNITY — 30.
▸ SPEC SHEET
Nemotron-4 340B — 340B Dense. ▸ SPECIFICATIONS
PARAMETERS 340B
ARCHITECTURE Dense Transformer
CONTEXT LENGTH 4K tokens
CAPABILITIES chat, reasoning
RELEASE DATE 2024-06-14
PROVIDER NVIDIA
FAMILY nemotron ▸ VRAM REQUIREMENTS
QUANT BPW VRAM QUALITY IQ2_XXS 2.38 101.6 GB 65% IQ2_M 2.93 125.0 GB 75% Q2_K 3.16 134.8 GB 78% IQ3_XXS 3.25 138.6 GB 82% IQ3_XS 3.5 149.2 GB 84% Q3_K_S 3.64 155.2 GB 85% IQ3_M 3.76 160.3 GB 86% Q3_K_M 4 170.5 GB 88% Q3_K_L 4.3 183.2 GB 90% IQ4_XS 4.46 190.0 GB 92% Q4_K_S 4.67 199.0 GB 93% Q4_K_M 4.89 208.3 GB 94% Q5_K_S 5.57 237.2 GB 96% Q5_K_M 5.7 242.7 GB 96% Q6_K 6.56 279.3 GB 97% Q8_0 8.5 361.7 GB 100% FP16 16 680.5 GB 100%
§ 01 BENCHMARK SCORES
HumanEval 73.2
MMLU-PRO 62.0
MATH 41.1
IFEval 85.1
Arena Elo 1482.0
§ 03 COMPATIBLE GPUs
5 @ Q4_K_M Feedback 