Claire 7B 0.1

Overview

LINAGORA's LoRA fine-tune of Falcon-7B specialized for spontaneous French dialogue. Released under CC-BY-NC-SA 4.0, with a separate Apache-licensed variant available for commercial work.

When to pick this model

• Research projects focused on conversational French
• Non-commercial prototypes needing native-feeling French dialogue
• Academic studies of spoken-style language modeling
• Baseline comparisons against modern French-capable LLMs

VRAM requirements by quantization

Quantization	VRAM required
Q4_K_M (recommended)	5 GB
Q5_K_M	6 GB
Q8_0	9 GB
FP16 (no quantization)	14 GB

VRAM figures include model weights plus a typical 8k KV cache and ~600 MB runtime overhead (Ollama / llama.cpp baseline). Add headroom for higher context lengths.

Strengths

• Natural, spoken-style French output
• Lightweight 7B footprint for local experimentation
• Backed by OpenLLM-France community work
• Targeted training on authentic French dialogue data

Limitations

• CC-BY-NC-SA license blocks most commercial use
• Tiny 2k context window by modern standards
• Built on aging Falcon-7B base
• Outclassed by Mistral and Qwen on general French tasks

Architecture & training

Architecture: LoRA fine-tune of Falcon-7B on spontaneous FR dialogue

Training: LINAGORA + OpenLLM-France.

Quick start

# HuggingFace : OpenLLM-France/Claire-7B-0.1 (Apache : Claire-7B-Apache-0.1)

Or use the open-source MCP server to query this model from Claude Desktop, Cursor, or any MCP-compatible client.