OLMo 3 32B

Overview

Allen AI's fully open dense 32B with Think and Instruct variants, releasing weights, data, and code under Apache 2.0. The transparency benchmark for 32B-class models.

When to pick this model

• Regulated industries that must audit training data
• Academic and reproducibility research at scale
• EU AI Act compliance requiring full traceability
• Apache-licensed commercial deployments
• Choosing between toggleable Think and Instruct modes

VRAM requirements by quantization

Quantization	VRAM required
Q4_K_M (recommended)	19 GB
Q5_K_M	23 GB
Q8_0	35 GB
FP16 (no quantization)	64 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

• Complete training transparency at 32B scale
• Apache 2.0 across weights, data, and code
• Think and Instruct variants for different workloads
• Strongest auditable model for AI Act compliance

Limitations

• Benchmarks trail closed-data 32B models
• 64K context lags top competitors
• Less polished than commercial-tuned alternatives

Architecture & training

Architecture: Dense 32B · 100% open (weights + data + code)

Training: Allen AI. Think and Instruct variants.

Quick start

ollama run olmo-3:32b

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