Back to AI app discovery
Alpie Core logo

Alpie Core

A 4-bit reasoning model with frontier-level performance

Developer Tools

Quick facts

Free plan
Yes
Pricing
Free: open-source model under Apache 2.0.
Platforms
Web app, API, Self-hosted

Pricing verified Oct 31, 2026 · Source

Key features of Alpie Core

  • 4-bit quantized reasoning 32B-parameter model in 4-bit NF4 precision needs about 16 GB VRAM, roughly 75 percent less memory than FP16 baselines.
  • 65K-token context window Handles 65K-token inputs with up to 16,384-token outputs for long documents and extended reasoning chains.
  • Streaming token output Generates responses token by token in real time, with vLLM recommended for high-throughput serving.
  • Open-source Apache 2.0 Weights and code are Apache 2.0 licensed, free for research and commercial use, with checkpoints on Hugging Face.

Verified Oct 31, 2026 · Source

Who Alpie Core is for

Best for

Developers and researchers who need an open, efficient reasoning model for coding, math, and science work.

Not ideal for

Users without a GPU or dev environment cannot run the 4-bit weights locally; self-hosting needs about 16 GB VRAM.

Screenshots

Alpie Core Homepage

About Alpie Core

Alpie Core is a 32B reasoning model trained, fine-tuned, and served entirely at 4-bit precision. Built with a reasoning-first design, it delivers strong performance in multi-step reasoning and coding while using a fraction of the compute of full-precision models. Alpie Core is open source, OpenAI-compatible, supports long context, and is available via Hugging Face, Ollama, and a hosted API for real-world use.

Ratings & reviews

Alpie Core alternatives

Alpie Core FAQ

Is Alpie Core free?

Yes. Alpie Core is open source under Apache 2.0, so the weights and code are free for research and commercial use. Checkpoints are published on Hugging Face.

What hardware do I need to run Alpie Core?

About 16 GB of VRAM. The 4-bit NF4 quantization cuts memory use by roughly 75 percent versus full precision, so it runs on GPUs with 16 to 24 GB. The repo recommends vLLM as the inference engine.

What is Alpie Core built on?

It is built on the DeepSeek-32B model family and fine-tuned with LoRA and QLoRA techniques in 4-bit precision.

Announcements

No announcements yet.

Community activity

Recent follows, shares, ratings, and collection saves for Alpie Core.

No community activity yet. Follow or share Alpie Core to get things started.

Comments (0)

Sign in to comment
What kind of project would you trust Alpie Core with first?Reply