TAAFT
Free mode
100% free
Freemium
Free Trial
Deals
Companies
169Pi
Models
Alpie Core

Alpie Core

By 169Pi
Alpie-Core is a 32B-parameter reasoning model from 169Pi, fine-tuned in 4-bit NF4 precision with double quantization and parameter-efficient training methods such as LoRA and QLoRA. The repository describes it as a resource-optimized model for reasoning, mathematics, coding, and science, with a 65K-token context window, streaming support, and an approximate 16GB memory footprint for deployment. 169Pi reports results including 81.28% on MMLU, 92.75% on GSM8K, and 57.8% on SWE-Bench Verified, and says the model was trained using 8 NVIDIA Hopper GPUs. It is released under the Apache 2.0 license.
Text Gen 3
Released: September 23, 2025

Overview

Alpie-Core is 169Pi’s first-generation 4-bit quantized reasoning model, built on the DeepSeek-32B family and optimized for efficient reasoning, mathematics, coding, and science tasks. It combines QLoRA, LoRA, and synthetic dataset distillation, supports a 65K-token context, and is designed to deliver strong reasoning performance with far lower memory and compute costs.

💬Chatting💻Conversational coding🧠AI inference

About 169Pi

Alpie is a free AI assistant built by 169Pi, an Indian AI research lab. It is powered by Alpie Core, a 32B parameter 4-bit quantized reasoning model that achieves competitive benchmarks against larger models at a fraction of the cost (~$3.50/1M tokens). The platform offers collaborative AI search, multi-agent workflows, and an API for developers across text, code, and reasoning tasks.

Industry: Artificial Intelligence
Company Size: 42
Location: Mumbai, IN
Website: alpie.ai
View Company Profile

Tools using Alpie Core

No tools found for this model yet.

Related Models

  • Gen 7
    o4 mini high
    o4-mini-high is the higher-accuracy profile of the mini line. It trades a bit of speed for stronger reasoning and coding, with long context, tool and function calling, streaming, and reliable JSON.
    Text
    Released 1y ago
  • Gen 6
    DeepSeek V3
    Text
    Released 1y ago
  • Gen 3
    Qianfan VL 70B
    Qianfan-VL 70B is Baidu’s large vision-language model on the Qianfan platform. It ingests images (docs, charts, screenshots, photos) with text and produces grounded answers, featuring strong OCR and layout understanding, long context, tool/function calling, streaming, and reliable JSON outputs for multimodal RAG and enterprise apps.
    📜OCR🖼️3D image generation🎬Video dubbing🔍Image recognition
    Text
    Released 8mo ago
Last updated: April 2, 2026
0 AIs selected
Clear selection
#
Name
Task