Falcon 3

What is Falcon 3?

Falcon 3 is TII's open-weights language model. Released in 2024-12, it ships in 4 sizes (1B, 3B, 7B, 10B) and is licensed Apache 2.0. The most popular use case is small efficient chat on consumer hardware.

VRAM and hardware

The smallest 1B size needs at least 1 GB VRAM; the largest 10B needs around 1 GB.

Size	Min VRAM (Q4_K_M)	Recommended VRAM	Tokens/sec on 3090
1B	1 GB	2 GB	~34 tok/s
3B	2 GB	3 GB	~31 tok/s
7B	5 GB	8 GB	~25 tok/s
10B	7 GB	11 GB	~23 tok/s

How to run Falcon 3 locally

Option 1: Ollama (simplest)

ollama pull falcon-3
ollama run falcon-3

Option 2: Mullama (production)

mullama pull falcon-3
mullama run falcon-3

Option 3: llama.cpp (CLI)

# Download a GGUF from Hugging Face (search "falcon-3 gguf")
./llama-cli -m falcon-3.Q4_K_M.gguf -p "Hello, AI!"

Option 4: Python with Mullama or llama-cpp-python

from mullama import Model, Context
model = Model.load("falcon-3.Q4_K_M.gguf", n_gpu_layers=99)
ctx = Context(model, n_ctx=4096)
print(ctx.generate("Hello, AI!", 256))

What you can build with Falcon 3

• chat
• multilingual
• edge

Hardware it fits on

• Apple Silicon (8GB): 1B in MLX
• Apple Silicon (16GB): 3B in MLX
• Apple Silicon (32GB): 7B in MLX or GGUF
• Apple Silicon (64GB+): 10B in MLX
• RTX 3090 (24GB): 7B at Q4_K_M
• RTX 4090 (24GB): same as 3090, ~30% faster
• RTX 5090 (32GB): 7B at Q8_0
• 2× RTX 6000 Ada (96GB total): 10B at Q4_K_M

Related models in the Falcon family

This is the only Falcon-family model in the catalog right now.

Other models you might consider

• Llama 3.1 (Meta) — 8B, general-purpose chat, agentic workflows, code
• Llama 3.2 (Meta) — 1B, low-end hardware, mobile, edge deployment
• Llama 3.2 Vision (Meta) — 11B, vision + chat on desktop GPUs
• Llama 3.3 (Meta) — 70B, best-in-class 70B for desktops
• Llama 4 (Meta) — Scout 17B (109B MoE), vision + long context (10M tokens)
• Qwen 2.5 (Alibaba) — 0.5B, strongest 7B-32B on consumer hardware