01.AIYiLlamaForCausalLM

Yi 6B Chat — Hardware Requirements & GPU Compatibility

Chat

Yi-6B-Chat is 01.AI's 6-billion-parameter bilingual (English/Chinese) chat model, instruction-tuned from the Yi-6B base model, part of the first-generation Yi series trained from scratch on a 3-trillion-token multilingual corpus. It uses the same Transformer structure popularized by Llama, though 01.AI states it is an independently trained model rather than a Llama derivative, and it was competitive with much larger contemporaries on benchmarks like the Hugging Face Open LLM Leaderboard and C-Eval at release. At 6B parameters it fits a single consumer GPU in half precision, and a much smaller card once 4-bit or 8-bit quantized. Context length is 4,096 tokens; a separate 200K-context variant of the base model is also available for longer documents. It is released under the Apache 2.0 license, permitting unrestricted commercial and research use, and was published in November 2023.

6.1B params · dense4K contextApache 2.0Nov 202320.6K downloads72 likes2.6K quant downloads

Your hardware

Verdicts below are for the reference RTX 4090. Set yours to personalise the page.

Detection reads your browser's GPU name with no data sent anywhere.

Runs great on a 24 GB GPU

Runs from 8 GB (Q4_K_M · 4K)Best on 24 GB (Q4_K_M · ~161 tok/s · 4K)Macs from 8 GB (Q4_K_M · ~18 tok/s)

Apache 2.0

Reference GPU · RTX 4090 24 GB

Estimates, ±20% · how we estimate

Run Yi 6B Chat

One recommended way. Every other method, OS and quant is in the step-by-step guide.Each way opens a step-by-step guide for your OS and quant.

Ways to runOther ways

each opens step-by-step

Run in cloud

Runs on RTX 3080 Ti (12 GB) (7 GB headroom) · Q4_K_M

Generation speed: ~146 tok/s; generation speed
Cost per 1M output tokens: $0.23; per 1M output tokens

Deployon Vast.aiBest valuefrom $0.12/hr

Compare GPUs →

Run Yi 6B Chat

4.2 GB VRAM at 4K context · fits the RTX 4090 (19.8 GB free)

Works on macOS, Windows and Linux.

Install Ollama (Linux)
$curl -fsSL https://ollama.com/install.sh | sh
Install Ollama (macOS / Windows)
Download and run the app; it installs the ollama command.
Download Ollama ↗
Run Q4_K_M
$ollama run hf.co/mradermacher/Yi-6B-Chat-GGUF:Q4_K_M
Raise the context window
Type it inside the chat. For the API server instead, start it with OLLAMA_CONTEXT_LENGTH=4096 ollama serve.
$/set parameter num_ctx 4096
A larger context needs more memory than the weights alone.
Use the API
Ollama serves its API on http://localhost:11434.

Works on macOS, Windows and Linux.

Install LM Studio
Desktop app for macOS (Apple silicon), Windows and Linux.
lmstudio.ai ↗
Download the model
In the app press Ctrl/Cmd + Shift + M and paste https://huggingface.co/mradermacher/Yi-6B-Chat-GGUF, then pick Q4_K_M. Or use the CLI (run LM Studio once first):
$lms get https://huggingface.co/mradermacher/Yi-6B-Chat-GGUF@Q4_K_M
Open on Hugging Face ↗
Load and chat
Open the model loader, select the model and set Context Length to 4096. Leave GPU offload at maximum if it fits.
Start the local server (optional)
$lms server start

Works on macOS, Windows and Linux.

Install llama.cpp (macOS)
MacPorts, Nix and conda-forge also work.
$brew install llama.cpp
Install options ↗
Install llama.cpp (Windows)
$winget install llama.cpp
Install llama.cpp (Linux)
Download the prebuilt llama-*-bin-ubuntu-* archive for your GPU backend (CPU, Vulkan) from the releases page, unpack it and run the binaries from that folder. Homebrew (brew install llama.cpp) and conda-forge also work on Linux.
llama.cpp releases ↗
Start the server
Downloads the model on first run. The chat UI and an OpenAI-compatible API are at http://localhost:8080.
$llama-server -hf mradermacher/Yi-6B-Chat-GGUF:Q4_K_M -c 4096 -ngl 99 --port 8080

Works on macOS, Windows and Linux.

Option A: Docker Model Runner
Docker Desktop: Settings > AI > Enable Docker Model Runner. Docker Engine on Linux: install the docker-model-plugin package.
$docker model run hf.co/mradermacher/Yi-6B-Chat-GGUF:Q4_K_M
Model Runner docs ↗
The quant tag after the repo is how Hugging Face cards write it; if your Docker version rejects it, drop the tag and pick the quant by repo.
Option B: llama.cpp server image (NVIDIA)
Needs the NVIDIA Container Toolkit. Models are cached in the llama-cache volume. Open http://localhost:8080.
$docker run --gpus all -p 8080:8080 -v llama-cache:/root/.cache ghcr.io/ggml-org/llama.cpp:server-cuda -hf mradermacher/Yi-6B-Chat-GGUF:Q4_K_M -c 4096 -ngl 99 --host 0.0.0.0 --port 8080
Image tags (cuda13, vulkan, rocm) ↗
Cache path inside the image is not in the docs; remove the -v flag if the volume complains.

Linux only.

Install vLLM
Linux only; on Windows run it inside WSL2. NVIDIA GPU shown; AMD ROCm, Intel and CPU builds have their own install pages.
$uv venv --python 3.12 --seed
$source .venv/bin/activate
$uv pip install vllm --torch-backend=auto
vLLM installation ↗
Serve the model
OpenAI-compatible API at http://localhost:8000/v1.
$vllm serve 01-ai/Yi-6B-Chat --max-model-len 4096
Runs the original weights, not the Q4_K_M GGUF; needs ~13 GB of GPU memory.

macOS only.

MLX supports only macOS, so follow these steps for that setup.

Install mlx-lm
$pip install mlx-lm
Chat in the terminal
$mlx_lm.chat --model mlx-community/Yi-6B-Chat-4bit
Downloads the model on first run. The MLX repo has its own quantisation; it may not match the quant selected above.
Or start a local server
Listens on http://localhost:8080.
$mlx_lm.server --model mlx-community/Yi-6B-Chat-4bit

Works on macOS, Windows and Linux.

Install Jan
jan.ai ↗
Find the model
Open the Hub tab, paste this model ID in Search, and choose Q4_K_M if offered.
$mradermacher/Yi-6B-Chat-GGUF
Use the local API (optional)
Settings > Local API Server > Start Server. It listens on http://127.0.0.1:1337.

Rent GPUs hourly to run Q4_K_M right there. Pricing comes from partner marketplaces and varies.

Available GPU rentals for Q4_K_M, prioritized by value
GPU	Speed	$/1M tokens	Rent
RTX 3080 Ti (12 GB)CheapestCheapest7 GB free	~146 tok/s	$0.23	Deploy on Vast.aifrom $0.12/hr (opens in new tab)	12 GB · 7 GB free · ~146 tok/s · $0.23 / 1M tokens
RTX 3060 12GB7 GB free	~58 tok/s	$0.30	Deploy on Vast.aifrom $0.061/hr (opens in new tab)	7 GB free · ~58 tok/s · $0.30 / 1M tokens
RTX 3060 Ti (8 GB)3 GB free	~72 tok/s	$0.30	Deploy on Vast.aifrom $0.076/hr (opens in new tab)	8 GB · 3 GB free · ~72 tok/s · $0.30 / 1M tokens
RTX 3070 (8 GB)3 GB free	~72 tok/s	$0.32	Deploy on Vast.aifrom $0.081/hr (opens in new tab)	8 GB · 3 GB free · ~72 tok/s · $0.32 / 1M tokens
RTX 4060 (8 GB)3 GB free	~43 tok/s	$0.55	Deploy on Vast.aifrom $0.086/hr (opens in new tab)	8 GB · 3 GB free · ~43 tok/s · $0.55 / 1M tokens
RTX 3080 (10 GB)5 GB free	~121 tok/s	$0.39	Deploy on RunPodfrom $0.17/hr (opens in new tab)	10 GB · 5 GB free · ~121 tok/s · $0.39 / 1M tokens
RTX 4000 Ada (20 GB)15 GB free	~58 tok/s	$0.97	Deploy on RunPodfrom $0.20/hr (opens in new tab)	20 GB · 15 GB free · ~58 tok/s · $0.97 / 1M tokens
RTX 3090 (24 GB)19 GB free	~150 tok/s	$0.41	Deploy on RunPodfrom $0.22/hr (opens in new tab)	24 GB · 19 GB free · ~150 tok/s · $0.41 / 1M tokens
RTX 4090 (24 GB)19 GB free	~161 tok/s	$0.59	Deploy on RunPodfrom $0.34/hr (opens in new tab)	24 GB · 19 GB free · ~161 tok/s · $0.59 / 1M tokens

Compare GPUs →

How Much VRAM Does Yi 6B Chat Need?

Select a quantization level; the listed hardware adheres to it.

VRAM	+32K	+128K	File	Quality	On RTX 4090 (reference)
2.1 GB	—	—	1.7 GB	2-bit · i-quant extreme compression — Importance-weighted 2-bit, extreme compression — significant quality loss	Fits: 4K context · ~312 tok/s
2.3 GB	—	—	1.9 GB	2-bit · i-quant extra small — Importance-weighted 2-bit, extra small	Fits: 4K context · ~291 tok/s
2.3 GB	—	—	2 GB	2-bit · i-quant small — Importance-weighted 2-bit, small	Fits: 4K context · ~281 tok/s
2.5 GB	—	—	2.2 GB	2-bit · i-quant medium — Importance-weighted 2-bit, medium	Fits: 4K context · ~264 tok/s
2.8 GB	—	—	2.4 GB	3-bit · i-quant — Importance-weighted 3-bit	Fits: 4K context · ~236 tok/s
2.9 GB	—	—	2.2 GB	2-bit · small K-quant — 2-bit small K-quant	Fits: 4K context · ~229 tok/s
2.9 GB	—	—	2.6 GB	3-bit · i-quant extra small — Importance-weighted 3-bit, extra small	Fits: 4K context · ~224 tok/s
3.0 GB	—	—	2.7 GB	3-bit · i-quant small — Importance-weighted 3-bit, small	Fits: 4K context · ~218 tok/s
3.0 GB	—	—	2.3 GB	2-bit · K-quant — 2-bit quantization with K-quant improvements	Fits: 4K context · ~218 tok/s
3.1 GB	—	—	2.7 GB	3-bit small — 3-bit small quantization	Fits: 4K context · ~212 tok/s
3.2 GB	—	—	2.8 GB	3-bit · i-quant medium — Importance-weighted 3-bit, medium	Fits: 4K context · ~207 tok/s
3.4 GB	—	—	3 GB	3-bit medium — 3-bit medium quantization	Fits: 4K context · ~193 tok/s
3.5 GB	—	—	3.5 GB	4-bit · legacy — 4-bit legacy quantization	Fits: 4K context · ~189 tok/s
3.5 GB	—	—	3.2 GB	3-bit large — 3-bit large quantization	Fits: 4K context · ~185 tok/s
3.7 GB	—	—	3.3 GB	4-bit · i-quant compact — Importance-weighted 4-bit, compact	Fits: 4K context · ~178 tok/s
3.8 GB	—	—	3.5 GB	4-bit small — 4-bit small quantization	Fits: 4K context · ~171 tok/s
3.8 GB	—	—	3.5 GB	4-bit · i-quant non-linear — Importance-weighted 4-bit, non-linear	Fits: 4K context · ~171 tok/s
3.8 GB	—	—	3.8 GB	4-bit · legacy+ — 4-bit legacy quantization with offset	Fits: 4K context · ~171 tok/s
4.1 GB	—	—	3.7 GB	4-bit medium — 4-bit medium quantization — most popular sweet spot	Fits: 4K context · ~161 tok/s
4.2 GB	—	—	4.2 GB	5-bit · legacy — 5-bit legacy quantization	Fits: 4K context · ~155 tok/s
4.6 GB	—	—	4.2 GB	5-bit small — 5-bit small quantization	Fits: 4K context · ~142 tok/s
4.8 GB	—	—	4.3 GB	5-bit medium — 5-bit medium quantization — good quality/size tradeoff	Fits: 4K context · ~138 tok/s
5.4 GB	—	—	5 GB	6-bit, very good — 6-bit quantization, very good quality	Fits: 4K context · ~121 tok/s
6.5 GB	—	—	6.4 GB	8-bit, near-lossless — 8-bit quantization, near-lossless	Fits: 4K context · ~101 tok/s
12.6 GB	—	—	~12.1 GB	16-bit · bfloat — Brain floating point 16 — preferred for training	Fits: 4K context · ~52 tok/s

est.calculated, no published file yet

Selected Q4_K_M

4-bit medium quantization

4.2GBat 4K context

Weights: 3.6 GB
KV cache: 0.3 GB
Overhead: 0.3 GB

Fits RTX 4090 (reference) · 19.8 GB free · ~161 tok/s

≥50 tok/s needs ≥313 GB/s · RTX 5090 ~286, RTX 3090 Ti ~161

Which Hardware Can Run Yi 6B Chat?

Q4_K_M · 4.1 GB

Yi 6B Chat (Q4_K_M) requires 4.1 GB of VRAM to load the model weights. For comfortable inference with headroom for KV cache and system overhead, 6+ GB is recommended. Using the full 4K context window can add up to 0.1 GB, bringing total usage to 4.2 GB. 52 GPUs can run it, including NVIDIA GeForce RTX 5090, NVIDIA GeForce RTX 3090 Ti. 163 devices with unified memory can run Yi 6B Chat, including Mac Studio (M3 Ultra, 256GB), Mac Studio (M3 Ultra, 512GB), MacBook Air 13" M3 (8 GB).

By VRAM tier · typical GPUs and Macs at Q4_K_M

Typical hardwareQuantVRAMtok/sContext that fits

8 GB · RTX 4060Q4_K_M · 4.2 GB · ~43 tok/s · 4K4.2~434K8 GB tier →

12 GB · RTX 3060 12GBQ4_K_M · 4.2 GB · ~58 tok/s · 4K4.2~584K12 GB tier →

16 GB · RTX 5060 Ti 16GBQ4_K_M · 4.2 GB · ~72 tok/s · 4K4.2~724K16 GB tier →

24 GB · RTX 4090referenceQ4_K_M · 4.2 GB · ~161 tok/s · 4K4.2~1614K24 GB tier →

Mac 32 GB · Mac Mini M4Q4_K_M · 4.2 GB · ~21 tok/s · 4K4.2~214KMacs →

Mac 64 GB · Mac Studio M4 MaxQ4_K_M · 4.2 GB · ~94 tok/s · 4K4.2~944KMacs →

How good is Yi 6B Chat?

llmrun Score blends public benchmark boards into one number; 100 = the average open model.

Not enough data yet for an llmrun Score.

A Score needs results on enough independent boards. How the Composite Score is built

Benchmarks

MMLU-Pro
TIGER-LabIndependent
28.8%
#123/146 open
Independent
MMLU
Epoch AI · Nov 2023Lab-reported
61.0%
#45/85 open
Lab-reported
BIG-Bench Hard
Epoch AI · Nov 2023Lab-reported
47.2%
#21/38 open
Lab-reported
GSM8K
Epoch AI · Nov 2023Lab-reported
44.9%
#35/62 open
Lab-reported

All 4 Yi 6B Chat benchmarks

Alternatives to Yi 6B Chat

Yi family · 6 models →

Similar models

6tok/s on RTX 4090

Model	Score	VRAM	tok/s	Why it is similar
Chatglm3 6BZ.ai	—	5.2 GB	~159	same size
TLive Omni 4BTaoLiveAIGC	—	13.3 GB	~54	same size
Z Image TurboTongyi-MAI	—	15.9 GB	~48	same size
GPT J 6BEleutherAI	—	4 GB	~166	same size
Ko PlatYi 6Bkyujinpy	—	4 GB	~163	same size
Yi 1.5 6B Chat01.AI	—	4.2 GB	~161	same size

Specifications

Publisher: 01.AI
Family: Yi
Parameters: 6.1B
Architecture: LlamaForCausalLM
Context Length: 4,096 tokens
Vocabulary Size: 64,000
Release Date: 2023-11-22
License: Apache 2.0

Resources

Frequently Asked Questions

How much VRAM does Yi 6B Chat need?

Yi 6B Chat requires 4.1 GB of VRAM at Q4_K_M, or 12.6 GB at BF16.

VRAM = Weights + KV Cache + Overhead

Weights = 6.1B × 4.8 bits ÷ 8 = 3.6 GB

KV Cache + Overhead ≈ 0.5 GB (at 2K context + ~0.3 GB framework)

Fit ratings and hardware model lists check this model with room for a 16K-token context, which needs a little more memory.

KV Cache + Overhead ≈ 0.6 GB (at full 4K context)

VRAM usage by quantization

Q4_K_M

4.1 GB

Q4_K_M + full context

4.2 GB

Learn more about VRAM estimation →

What's the best quantization for Yi 6B Chat?

For Yi 6B Chat, Q4_K_M (4.1 GB) offers the best balance of quality and VRAM usage. Q5_0 (4.2 GB) provides better quality if you have the VRAM. The smallest option is IQ2_XXS at 2.1 GB.

VRAM requirement by quantization

IQ2_XXS

2.1 GB~53%

IQ3_XS

2.9 GB~73%

Q4_0

3.5 GB~85%

Q4_K_M ★

4.1 GB~89%

Q5_0

4.2 GB~90%

BF16

12.6 GB~100%

★ Recommended — best balance of quality and VRAM usage.

Learn more about quantization →

Can I run Yi 6B Chat on a Mac?

Yes — MacBook Air 13" M3 (8 GB) and 145 other Macs can run Yi 6B Chat. Apple Silicon uses unified memory, so the model shares RAM with the system. At Q4_K_M you need at least 4.1 GB of usable unified memory (RAM minus macOS overhead).

Can I run Yi 6B Chat locally?

Yes — Yi 6B Chat can run locally on consumer hardware. Popular tools include Ollama, LM Studio, and llama.cpp.

How fast is Yi 6B Chat?

At Q4_K_M, Yi 6B Chat can reach ~1179 tok/s on AMD Instinct MI350X. On NVIDIA GeForce RTX 4090: ~161 tok/s. Speed depends mainly on GPU memory bandwidth. Real-world results typically within ±20%.

tok/s = (bandwidth GB/s ÷ model GB) × efficiency

Example: NVIDIA B200 → 8000 ÷ 4.07 × 0.65 = ~1278 tok/s

Estimated speed at Q4_K_M (4.1 GB)

NVIDIA B200

~1278 tok/s

NVIDIA GeForce RTX 4090

~161 tok/s

NVIDIA B300

~1278 tok/s

AMD Instinct MI350X

~1179 tok/s

Real-world results typically within ±20%. Speed depends on batch size, quantization kernel, and software stack.

Learn more about tok/s estimation →

What's the download size of Yi 6B Chat?

At Q4_K_M, the download is about 3.64 GB. The full-precision BF16 version is 12.12 GB. The smallest option (IQ2_XXS) is 1.67 GB.

Which GPUs can run Yi 6B Chat?

52 consumer GPUs can run Yi 6B Chat at Q4_K_M. Top options include AMD Radeon RX 6700 XT, AMD Radeon RX 6800, AMD Radeon RX 6800 XT. 52 GPUs have plenty of headroom for comfortable inference.

Which devices can run Yi 6B Chat?

163 devices with unified memory can run Yi 6B Chat at Q4_K_M (4.1 GB), including AMD Ryzen AI 9 HX 370 (Strix Point) Laptop, ASUS Ascent GX10, Beelink GTR9 Pro (Ryzen AI Max+ 395, 128 GB), Framework Desktop (Ryzen AI Max+ 395, 128 GB). Apple Silicon Macs use unified memory shared between CPU and GPU, making them well-suited for local LLM inference.