Aya Expanse 8B

Overview

Cohere For AI's multilingual 8B covering 23 languages, outperforming Gemma 2 9B and Llama 3.1 8B in its language set. CC-BY-NC — non-commercial only.

When to pick this model

• You're doing multilingual research that doesn't require commercial use
• You need strong coverage of low-resource languages at the 8B tier
• You're benchmarking against Gemma 2 9B and Llama 3.1 8B on non-English tasks
• You're building an internal evaluation harness

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)	16 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.

Published benchmark scores

Benchmark	Score
Dolly (vs Llama 3.1 8B)	83.9

Scores published by the model author or aggregated from public leaderboards. Re-measured monthly by our editorial team.

Strengths

• 23 language coverage with strong low-resource performance
• Beats Gemma 2 9B and Llama 3.1 8B on multilingual benchmarks
• Particularly strong on low-resource languages
• Compact 8B footprint

Limitations

• CC-BY-NC 4.0 — no commercial deployment
• Only 8K context
• Outclassed by Qwen 3 8B on most general tasks

Architecture & training

Architecture: Dense · 32 layers · 32 heads · SwiGLU · GQA · SentencePiece ~128k vocab

Training: 23 languages, multilingual focus.

Quick start

ollama run aya-expanse:8b

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