The real cost to run Ouro 2.6B Thinking

ByteDance · 2.7B parameters · 65.5K context · commercial OK

Ouro 2.6B Thinking — 2.7B params (ByteDance). Auto-indexed from the Hugging Face Hub (15,718 downloads). Parameter count is exact; download size and quantizations are estimates.

What it costs to run Ouro 2.6B Thinking — $0 markup

On your own machine — $0. Runs free locally if you have about 10 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 — $0.12/1M tokens (blended input + output). No GPU to manage (rough estimate for this size).

Key facts

Parameters	2.7B
Context window	65.5K tokens
Recommended quant	Q4_K_M
VRAM to run	~10 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

Ouro 1.4B — 1.4B, ByteDance
gpt neo 2.7B — 2.7B, EleutherAI
phi 2 — 2.8B, microsoft
stablelm 3b 4e1t — 2.8B, stabilityai
gemma 2 2b it — 2.6B, google

Browse: All models · Compare