CoolFace
Modelpublic

immortaltatsu/ghost-ai-pro-gguf

sourceHugging Faceupdated 5mo agoView on Hugging Face
0likes46downloads
Model Card

Ghost AI Pro

On-device financial intelligence — built for Ghost AI on Solana.

Ghost AI Pro is part of Ghost AI, an on-device privacy-first crypto application running on Solana. The model powers private, local AI features inside the Ghost AI app — your prompts, wallet context, and financial data never leave your device. Ghost AI Pro is a fork of LiquidAI's LFM 2.5 (1.2B Instruct), rebranded and packaged for on-device financial use cases. All underlying model weights, architecture, and capabilities derive from LFM 2.5 — credit and licensing terms remain with the upstream project.

Ghost AI Pro is a compact, quantized language model engineered to run privately on consumer hardware — phones, laptops, edge devices — without sending a single token to the cloud. It is purpose-tuned for the financial domain: portfolio reasoning, market commentary, statement parsing, expense categorization, on-chain activity summarization, and conversational assistance for personal, professional, and crypto-native finance workflows.

For ultra-light deployments (mobile, embedded), see the smaller [Ghost AI](https://huggingface.co/immortaltatsu/ghost-ai-gguf) (350M, Q6_K, ~280MB).

Ghost AI model family

TierRepoBaseQuantizationSizeBest for
Ghost AIimmortaltatsu/ghost-ai-ggufLFM 2.5 350MQ6_K~280 MBMobile, embedded, instant responses
Ghost AI Pro (this)immortaltatsu/ghost-ai-pro-ggufLFM 2.5 1.2B InstructQ4KS~700 MBDesktop, deeper reasoning

Why Ghost AI Pro

  • —On-device, always private. Runs locally via llama.cpp, LM Studio, Ollama, or any GGUF-compatible runtime. Your prompts, wallet data, and financial context never leave your device.
  • —Built for finance + crypto. Optimized prompting and behavior around budgeting, markets, accounting concepts, on-chain transactions, and personal-finance reasoning.
  • —Native to Ghost AI on Solana. Powers the local AI surface of the Ghost AI app — a privacy-first crypto experience on the Solana blockchain.
  • —Lean footprint. Q4KS quantization — ~700MB on disk, comfortable on 4GB+ RAM devices.
  • —Fast inference. Designed for real-time, low-latency interaction on CPU and consumer GPU.

Files

FileQuantizationSizeNotes
ghost-ai-pro-Q4_K_S.ggufQ4KS~700 MBRecommended default. Best balance of quality and footprint for on-device use.

Usage

llama.cpp

bash
./llama-cli -m ghost-ai-pro-Q4_K_S.gguf -p "Summarize this quarter's spending: ..." -n 256

Ollama (Modelfile)

FROM ./ghost-ai-pro-Q4_K_S.gguf
PARAMETER temperature 0.7
PARAMETER num_ctx 4096
bash
ollama create ghost-ai-pro -f Modelfile
ollama run ghost-ai-pro

LM Studio

Download the .gguf file and load it directly from the LM Studio model browser.

Intended Use

Ghost AI Pro is intended for:

  • —Powering the on-device AI layer of the Ghost AI crypto app on Solana
  • —Private wallet, transaction, and portfolio explanations
  • —Personal finance assistants and budgeting copilots
  • —On-device document Q&A over statements, invoices, and reports
  • —Lightweight market, DeFi, and economic commentary generation
  • —Embedded assistants in fintech and crypto apps where privacy is non-negotiable

Limitations

Ghost AI Pro is a small model. It is not a substitute for licensed financial advice, accounting, or legal counsel, and it is not a trading or transaction-signing agent. Verify any numerical reasoning, dates, ticker symbols, token addresses, on-chain claims, and regulatory statements against authoritative sources before acting on them. Always confirm wallet operations through the Ghost AI app's signing flow — never from raw model output.

About Ghost AI

Ghost AI is an on-device, privacy-first crypto app built on Solana. It brings local AI directly into the wallet experience — letting users explore transactions, balances, and DeFi context with a model that runs entirely on their own device. Nothing is logged, sent to a server, or shared with a third party.

Ghost AI Pro is the model that powers this on-device intelligence layer.

Upstream

Ghost AI Pro is a fork of [LiquidAI/LFM2.5-1.2B-Instruct](https://huggingface.co/LiquidAI/LFM2.5-1.2B-Instruct). The GGUF quantizations are sourced from [unsloth/LFM2.5-1.2B-Instruct-GGUF](https://huggingface.co/unsloth/LFM2.5-1.2B-Instruct-GGUF).

License

Released for research and on-device application use. This model is governed by the upstream LFM Open License issued by Liquid AI — please review the original terms before redistribution or commercial deployment.

Ghost AI Pro — intelligence that stays with you.