# Hugging Face — Falcon-H1: A Family of Hybrid-Head Language Models Redefining Efficiency and Performance

- Company: Hugging Face (huggingface.co)
- Announced: 2025-05-21T06:52:13+00:00
- Category: open-weight-release
- Subject: Platform
- Open weights: yes
- Models affected: Falcon-H1-0.5B-Base, Falcon-H1-0.5B-Instruct, Falcon-H1-1.5B-Base, Falcon-H1-1.5B-Instruct, Falcon-H1-1.5B-Deep-Base, Falcon-H1-1.5B-Deep-Instruct, Falcon-H1-3B-Base, Falcon-H1-3B-Instruct, Falcon-H1-7B-Base, Falcon-H1-7B-Instruct, Falcon-H1-34B-Base, Falcon-H1-34B-Instruct
- License: Apache 2.0
- Context window: 256K context length
- Source: https://huggingface.co/blog/tiiuae/falcon-h1
- Record: https://forck.live/items/1691-falcon-h1-a-family-of-hybrid-head-language-models-redefining-efficiency-and

TII UAE announces the Falcon-H1 series, a family of six open-source hybrid-head language models ranging from 0.5B to 34B parameters, with base and instruction-tuned variants. The models combine attention and Mamba-2 SSM heads, support up to 256K context length, and are released under Apache 2.0.

## Evidence

Verbatim from https://huggingface.co/blog/tiiuae/falcon-h1:

> Today, we are proud to introduce the Falcon-H1 series, a collection of six open-source models ranging from 0.5B to 34B parameters, each available in both base and instruction-tuned variants.

---

Record: https://forck.live/items/1691-falcon-h1-a-family-of-hybrid-head-language-models-redefining-efficiency-and
Catalogue: https://forck.live/llms.txt
Feed: https://forck.live/feed.md