Provider pricing

Lamini AI API pricing

Lamini is an enterprise LLM platform focused on fine-tuning open-source models with its Memory Tuning technique to cut hallucinations on factual recall. Customers run Lamini in their own VPC or use the hosted cloud, with per-token inference and per-job tuning pricing.

Founded2022
HeadquartersPalo Alto, CA
Tracked modelsDirectory

Best for

Enterprises fine-tuning open-source LLMs on proprietary knowledge bases.

Strengths

  • Memory Tuning reduces hallucinations on private data
  • Runs in customer VPC on AMD or NVIDIA GPUs
  • Fine-tunes Llama, Mistral, and Qwen with one API call
  • Pay per tuning job + per-token inference