6.9B
- Released
- 2y ago
- Size
- 272Mparameters
Smallest OpenELM language model, released with its full training and inference framework, in base and instruct versions.
How it compares
Frequently asked questions
Running locally
Estimated VRAM is 0.7 GB at Q4, 0.8 GB at Q8 and 1.1 GB at FP16. These are estimates from the 272M parameter count, covering weights plus about 15% and 0.5 GB for a short context.
Yes. A 24 GB, 16 GB, 12 GB or 8 GB GPU fits OpenELM 270M at FP16, estimated at 1.1 GB.
Yes. Q4 is estimated at 0.7 GB, and assuming macOS lets the GPU use about 75% of unified memory, the smallest common Mac memory size that fits Q4 is 8 GB.
Privacy and data
Yes. OpenELM 270M has open weights, so you can self-host it for full control of your data.
Capabilities and specs
No. OpenELM 270M accepts text input only.
Quality and benchmarks
Matched on model kind and release date, with no comparable price, the closest alternatives are OLMo 1.7 7B by Ai2, Hermes 2 Pro Llama 3 8B by Nous and Mistral 7B Instruct v0.3 by Mistral AI.
OLMo 1.7 7BOpenELM 270M vs OLMo 1.7 7BHermes 2 Pro Llama 3 8BOpenELM 270M vs Hermes 2 Pro Llama 3 8BMistral 7B Instruct v0.3OpenELM 270M vs Mistral 7B Instruct v0.3
Licensing and access
Open weights. The license is apple-amlr.
The weights are available on Hugging Face in the apple/OpenELM-270M repository.
Basics
OpenELM 270M was made by Apple and released on April 23, 2024.
Yes. OpenELM 270M has open weights.
Generated from Vertical Tabs data, updated November 3, 2026.
Discussion
No comments yet
Be the first to share your thoughts.