The real cost to run MiniMax M3 EAGLE3

Inferact · 3.3B parameters · 1M context · commercial OK

MiniMax M3 EAGLE3 — 3.3B params (Inferact). Auto-indexed from the Hugging Face Hub (10,467 downloads). Parameter count is exact; download size and quantizations are estimates.

What it costs to run MiniMax M3 EAGLE3 — $0 markup

On your own machine — $0. Runs free locally if you have about 4 GB of VRAM at Q4_K_M (point Spanvero at LM Studio, Ollama, or llama.cpp).
On your own rented GPU — from $0.26/hr. NVIDIA RTX 3090 24GB at the direct Vast.ai price (≈ $0.04 a session), $0 markup.
Via your own API key — $0.13/1M tokens (blended input + output). No GPU to manage (rough estimate for this size).

Key facts

Parameters	3.3B
Context window	1M tokens
Recommended quant	Q4_K_M
VRAM to run	~4 GB (at Q4_K_M, 16.4K context)
Download size	~2 GB
License	Commercial use OK

Open the free Spanvero advisor → for the live, interactive math for your exact workload and hardware.

Related models

tiny aya base — 3.3B, CohereLabs
tiny aya global — 3.3B, CohereLabs
Llama 3.2 3B — 3.2B, meta-llama
Llama 3.2 3B Instruct — 3.2B, unsloth
Llama 3.2 3B Instruct pythonic — 3.2B, baseten

Browse: All models · Compare