LLaVA 1.6 34B on NVIDIA GeForce RTX 4080 SUPER

Can the NVIDIA GeForce RTX 4080 SUPER run LLaVA 1.6 34B?

Yes, it runs well (offloads to RAM)

Runs well (offloads to RAM). At Q2_K it needs about 16.3 GB of 16.0 GB usable, with an estimated speed of 35.0 tok/s.

Best quant: Q2_K
Generation speed: 35.0 tok/s
Prompt processing: 1,065 tok/s
First token (1K prompt): 1.0 s
Largest safe context: 2K

Best app to run it on the NVIDIA GeForce RTX 4080 SUPER

On Windows

Ollama

Command line + local API · runs on llama.cpp (ggml), MLX (Apple silicon, preview)

Splits the model between GPU memory and system RAM, which this setup needs.
One command with Ollama’s own build of this model.
Serves an OpenAI-compatible API that other apps and coding tools can use.

ollama run llava:34b-v1.6-q2_K

Docker Model Runner Splits the model between GPU memory and system RAM, which this setup needs.
Text Generation WebUI Splits the model between GPU memory and system RAM, which this setup needs.

On Linux

Ollama

Command line + local API · runs on llama.cpp (ggml), MLX (Apple silicon, preview)

Splits the model between GPU memory and system RAM, which this setup needs.
One command with Ollama’s own build of this model.
Serves an OpenAI-compatible API that other apps and coding tools can use.

ollama run llava:34b-v1.6-q2_K

Docker Model Runner Splits the model between GPU memory and system RAM, which this setup needs.
Text Generation WebUI Splits the model between GPU memory and system RAM, which this setup needs.

How we choose: the files published for this model, your operating system and GPU, whether it fits fully in memory, and published speed tests (MLX vs llama.cpp on Apple silicon, vLLM vs llama.cpp vs Ollama on NVIDIA). Details.

Download LLaVA 1.6 34B

Recommended for the NVIDIA GeForce RTX 4080 SUPER: Q2_K. Every other quantization has its own download link in the table below.

Quickest · Ollama

One command downloads a ready-to-run version and starts a chat.

Get Ollama at ollama.com/download (opens in a new tab) · model page (opens in a new tab)

ollama run llava:34b-v1.6-q2_K

Original weights · Hugging Face

Full-precision files from LLaVA, for fine-tuning, vLLM, Transformers.

huggingface.co/llava-hf/llava-v1.6-34b-hf (opens in a new tab)

No sign-in needed.

Links checked 3 days ago against Hugging Face, Ollama and GitHub.

How it’s rated

Quality 30Basic quality (D) · 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking.: How capable the model itself is at Q2_K, from its editorial quality tier scaled by the quantization or precision. It’s the same on any hardware that runs this variant. Scale: A+ 85+ · A 70+ · B 55+ · C 40+ · D under 40.
Runs Runs welloffloads to RAM: How well it runs on the NVIDIA GeForce RTX 4080 SUPER: speed, capped by how well it fits in memory. Runs great: 30+ tokens/s (or near the image, video or decision speed target) · Runs well: 10+ tokens/s · Slow: 3+ tokens/s · Very slow: under 3 · Won’t run: doesn’t fit even with offload. A tight fit or RAM offload caps it at Runs well; heavy offload caps it at Slow.

Match score

To order lists, each result also gets a match score out of 100 from four parts. It favors the most capable model that still runs well.

Model qualityEditorial quality tier of the model, scaled by the quantization or precision used.	13.9 / 45
Memory fitFull marks when everything sits in GPU memory; fewer when part spills to system RAM.	10.0 / 25
SpeedGeneration speed against a 40 tokens/s target for LLMs, or render time against a target for image and video.	14.0 / 15
Memory useRewards using the memory you have, so a bigger, better variant beats a tiny one that leaves most memory idle.	15.0 / 15
Match score	52.9 / 100

The match score orders results; it is not shown as a grade. Quality and Runs are the two ratings in the tables.

Every quantization on the NVIDIA GeForce RTX 4080 SUPER

Quant	Memory needed	Available	Speed	Runs	Quality	Download
`Q2_K`	16.3 GB	16.0 GB	35.0 tok/s	Runs welloffloads to RAM	30Basic quality (D) · 30 of 100: how capable the model is at Q2_K, whatever it runs on. Match score 52/100 (quality, fit, speed and memory use) sets the ranking.
`Q3_K_M`	19.4 GB	16.0 GB	11.1 tok/s	Runs welloffloads to RAM	36Basic quality (D) · 36 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 49/100 (quality, fit, speed and memory use) sets the ranking.
`Q4_K_M`	23.6 GB	16.0 GB	5.6 tok/s	Slowheavy RAM offload	41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 43/100 (quality, fit, speed and memory use) sets the ranking.
`Q5_K_M`	27.1 GB	16.0 GB	3.9 tok/s	Slowheavy RAM offload	42Fair quality (C) · 42 of 100: how capable the model is at Q5_K_M, whatever it runs on. Match score 42/100 (quality, fit, speed and memory use) sets the ranking.
`Q6_K`	30.9 GB	16.0 GB	2.9 tok/s	Very slowheavy RAM offload	43Fair quality (C) · 43 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 42/100 (quality, fit, speed and memory use) sets the ranking.
`Q8_0`	39.2 GB	16.0 GB	—	Won't run	44Fair quality (C) · 44 of 100: how capable the model is at Q8_0, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking.
`F16`	71.9 GB	16.0 GB	—	Won't run	44Fair quality (C) · 44 of 100: how capable the model is at F16, whatever it runs on. Match score 5/100 (quality, fit, speed and memory use) sets the ranking.	Original weights (opens in a new tab)

Memory figures assume 4K of context. Longer conversations or documents grow the KV cache.

Other hardware that runs it

41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 73/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 5090 Q4_K_M · 63.3 tok/s · Runs great
36Basic quality (D) · 36 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 3090 Q3_K_M · 42.7 tok/s · Runs great
36Basic quality (D) · 36 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA GeForce RTX 4090 Q3_K_M · 46.0 tok/s · Runs great
36Basic quality (D) · 36 of 100: how capable the model is at Q3_K_M, whatever it runs on. Match score 71/100 (quality, fit, speed and memory use) sets the ranking. AMD Radeon RX 7900 XTX Q3_K_M · 37.8 tok/s · Runs great
43Fair quality (C) · 43 of 100: how capable the model is at Q6_K, whatever it runs on. Match score 67/100 (quality, fit, speed and memory use) sets the ranking. NVIDIA RTX PRO 6000 Blackwell Q6_K · 43.9 tok/s · Runs great
41Fair quality (C) · 41 of 100: how capable the model is at Q4_K_M, whatever it runs on. Match score 66/100 (quality, fit, speed and memory use) sets the ranking. Apple M4 Max (64 GB) Q4_K_M · 16.6 tok/s · Runs well

Quick answers

How much memory does LLaVA 1.6 34B need on the RTX 4080 SUPER 16GB?

At Q2_K it needs about 16.3 GB. The RTX 4080 SUPER 16GB has 16.0 GB usable.

How fast is LLaVA 1.6 34B on the RTX 4080 SUPER 16GB?

About 35.0 tok/s at Q2_K, an estimate from the hardware’s memory bandwidth and the size of the model.

Which quantization of LLaVA 1.6 34B should I use on the RTX 4080 SUPER 16GB?

Q2_K. Of the versions that fit, it is the one that runs best; when two run equally well, the check prefers the higher-rated one.

Estimates come from published hardware specs and our open formulas, not from running the model on your machine. Real speed varies with runtime, drivers and settings. Numbers marked “Measured” come from a linked benchmark.