Tested
Tested on the exact GPU, model digest, runtime and context shown.
Local AI compatibility
Know which local AI models your NVIDIA GPU can actually run.
Match your NVIDIA hardware with exact Ollama models, or check the hardware an Ollama model needs. Physical tests stay clearly separate from estimates.
✓ Exact model variants ✓ Offload explained ✓ No made-up tok/s
700
Canonical models
574/700
FULL compatibility verdicts
22
Reference hardware profiles
1007
Quantizations tracked
Hardware → models
What can I run?
Select what is in your machine. Results update in seconds.
Model → hardware
What hardware does it need?
Compare one exact Ollama variant across known NVIDIA configurations.
Before → after
What would this upgrade unlock?
See which model compatibility results improve with the new setup.
Trust model
Three labels. No blurred lines.
Estimated
A conservative memory precheck. Useful, but not a physical benchmark.
Not enough evidence
Test data is incomplete or the combination sits too close to a boundary.
Explore
Browse the full catalog
Every canonical model carries an explicit FULL, PARTIAL or BASIC compatibility verdict — never a guess dressed up as a fact.