Molmo 72B
By Allen AI · United States
Updated 2026-07-13
Overview
Allen AI's flagship Apache 2.0 VLM built on Qwen2-72B, ranked #2 in human evaluation behind only GPT-4o for visual understanding.
When to pick this model
- On-prem replacement for GPT-4o vision in regulated environments
- High-stakes visual analysis where quality dominates cost
- Research benchmarks demanding open weights at frontier quality
- Document and diagram understanding at scale
- Multi-GPU deployments already provisioned for 70B-class models
VRAM requirements by quantization
| Quantization | VRAM required |
|---|---|
| Q4_K_M (recommended) | 42 GB |
| Q5_K_M | 50 GB |
| Q8_0 | 78 GB |
| FP16 (no quantization) | 144 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.
In practice, Molmo 72B spills past single consumer GPUs even at Q4_K_M (42 GB) — think dual-GPU or workstation cards. Stepping up to Q8_0 nearly doubles the footprint to 78 GB, and unquantized FP16 weights take 144 GB — plan your GPU around the Q4 or Q5 figure unless you specifically need the higher fidelity.
Without a GPU, Molmo 72B needs roughly 64 GB of system RAM to run on CPU via llama.cpp or Ollama — workable for background jobs, but far slower than GPU inference. Throughput estimates from our compatibility engine: around 1 tokens/sec on entry-level GPUs, on the order of 5 tokens/sec on a mid-range card, and up to 18 tokens/sec on high-end hardware — assuming the chosen quantization fully fits in VRAM.
What hardware do you need
The table below matches Molmo 72B to common GPU memory tiers, using the highest-fidelity quantization that fully fits each card class. Spilling layers to system RAM works but costs most of the speed, so size your card to the quantization you actually want to run.
| GPU memory | Example cards | Best fit for Molmo 72B |
|---|---|---|
| 8 GB | RTX 5070 Laptop, RTX 5060, RTX 5060 Ti 8GB | Does not fit — needs 42 GB at Q4_K_M |
| 12 GB | RTX 5070, RTX 5070 Ti Laptop, RTX 4080 Laptop | Does not fit — needs 42 GB at Q4_K_M |
| 16 GB | RTX 5080, RTX 4080 Super, Radeon RX 9070 XT | Does not fit — needs 42 GB at Q4_K_M |
| 24 GB | RTX 4090, Radeon RX 7900 XTX, RTX 5090 Laptop | Does not fit — needs 42 GB at Q4_K_M |
| 32 GB | RTX 5090 | Does not fit — needs 42 GB at Q4_K_M |
Which GPU should you buy to run Molmo 72B?
To run Molmo 72B locally at Q4, you need ~42 GB of VRAM. The best value for this is a Apple Mac Studio (64+ GB unified memory).
As an Amazon Associate, BestLLMfor earns from qualifying purchases, at no extra cost to you. It does not influence our independent rankings.
Published benchmark scores
| Benchmark | Score |
|---|---|
| MMMU | 72.2 |
Scores published by the model author or aggregated from public leaderboards. Re-measured monthly by our editorial team.
To put Molmo 72B in context: its MMMU score of 72.2 ranks #1 of the 8 catalog models with a published MMMU result (catalog median 62.8). Scores are author-published and measured under different harnesses, so treat gaps of a few points as noise rather than a verdict.
Strengths
- Top-tier vision quality among open-weight VLMs
- Apache 2.0 license with PixMo open training data
- Strong on complex visual reasoning and dense scenes
- Human evaluation second only to GPT-4o
Limitations
- ~42 GB VRAM at Q4 typically requires 2-3 GPUs
- 4096-token context constrains long multimodal sessions
- No official GGUF release complicates llama.cpp use
Typical workloads
In our catalog grid, Molmo 72B is filed under Advanced Vision, VLM Research — the use cases where its size/quality trade-off makes the most sense. Its tags translate to concrete workloads: vision-language work — screenshots, charts, scanned documents.
Note the 4k-token context window — fine for short interactions, limiting for long documents or big retrieval contexts. The Apache 2.0 license is permissive, so shipping it inside a commercial product raises no special legal questions.
Architecture & training
Architecture: Dense · 72B vision · based on Qwen2 72B + OpenAI CLIP encoder
Training: AllenAI PixMo dataset, maximal 72B version.
The highest-quality fully open VLM — choose it when you have the GPUs and need GPT-4o-class vision on-prem.
Quick start
ollama run molmo:72bOr use the open-source MCP server to query this model from Claude Desktop, Cursor, or any MCP-compatible client.
Similar models worth comparing
Frequently asked questions
How much VRAM does Molmo 72B need?
At the recommended Q4_K_M quantization, Molmo 72B needs about 42 GB of VRAM. Q8_0 takes 78 GB, and unquantized FP16 weights take 144 GB.
Can Molmo 72B run without a GPU?
Yes — with roughly 64 GB of system RAM it runs CPU-only through llama.cpp or Ollama. Expect a fraction of GPU speed, which is fine for background or batch jobs but slow for interactive chat.
What context window does Molmo 72B support?
Molmo 72B supports a 4k-token context window (4,096 tokens).
Can I use Molmo 72B commercially?
Yes. Molmo 72B is released under Apache 2.0, a permissive open-source license that allows commercial use, modification and redistribution.
How fast is Molmo 72B on consumer hardware?
Our compatibility engine estimates on the order of 5 tokens/sec on a mid-range GPU and up to 18 tokens/sec on high-end cards, assuming the quantization fully fits in VRAM.
Which quantization of Molmo 72B should I download first?
Start with Q4_K_M (42 GB) — the standard size/quality sweet spot. Step up to Q5_K_M or Q8_0 only if you have VRAM headroom. It does not fit a single 24 GB consumer card — plan for multi-GPU or server hardware.