Pricing

One-time purchase. All quant variants + Ollama Modelfile included with every model.

Scout Tier

Consumer GPUs (8-24GB VRAM)

Lean-Agent-4B lightweight, fast tool calling

~4-5 GB $20
Lean-Agent-8B balanced agent model

~6-8 GB $20
Lean-Coder-8B code generation specialist

~6-8 GB $20
Lean-Agent-14B best quality in Scout tier

~10-12 GB $20

Deep Tier

Multi-GPU / high-VRAM (24GB+)

Lean-Agent-32B complex reasoning

~20-22 GB $25
Lean-Reason-32B extended chain-of-thought

~24-26 GB $25
Lean-Agent-30B-MoE fast MoE (3B active)

~18-20 GB $25

Bundles

Scout Bundle

All 4 Scout tier models

$45

save $35 vs. individual

Full Collection

All 7 models, both tiers

$75

save $80 vs. individual

Subscription

All models + new releases

$15 /mo

always up to date

Every purchase includes

Model weights in GGUF format (Q4_K_M through Q8_0 variants)
Tuned Ollama Modelfile for drop-in use
Documentation for Ollama and llama.cpp setup
Commercial use — no restrictions on deployment