Start with memory fit, then check the model's task and setup instructions. A verdict estimates compatibility; it does not measure answer quality or speed.
Scenario assumptions
Illustrative 8-core CPU and 500 GB free disk, so these estimates do not check your actual CPU or storage. Available RAM, GPU budget, free VRAM and driver version are unknown. The engine applies memory reserves. GPU paths assume a working supported driver/backend; runtime-version compatibility is not checked. Apple memory is one shared pool. One GPU is evaluated; memory is not pooled across cards.
This configuration is not saved as your PC, and does not add to the completed-scan counter.
Four builds to compare by task
Editorial examples chosen for their documented tasks, all Q4_K_M at 4,096 tokens. Check the verdict before downloading. These are estimates, not benchmark rankings.
After opening Ollama's prompt, enter /set parameter num_ctx 4096 to match this calculation, within the model's supported maximum. Install a current Ollama release that supports the build. Setup and troubleshooting.
Showing up to 20 builds, fits first and smallest memory first. These technical defaults include base and embedding models. They are not a recommendation ranking. A model's documented maximum context still applies.
Entire model runs from VRAM; no offload to system RAM required.
Storage: Pass
500.0 GB free, 44 MB required.
Acceleration: Pass
cuda acceleration available.
Assumptions and missing information
KV cache size is estimated using a fallback formula because architecture metadata is incomplete. Actual memory may differ significantly.
The NVIDIA driver version was not measured for every CUDA GPU. Ollama requires driver 550+, or 570+ for compute capability 5.0–6.2; update the driver before relying on GPU acceleration.
This is a hypothetical PC. No inference benchmark was run.
GB uses 1,024³ bytes. Memory includes weights, context cache and runtime overhead.
Entire model runs from VRAM; no offload to system RAM required.
Storage: Pass
500.0 GB free, 44 MB required.
Acceleration: Pass
cuda acceleration available.
Assumptions and missing information
KV cache size is estimated using a fallback formula because architecture metadata is incomplete. Actual memory may differ significantly.
The NVIDIA driver version was not measured for every CUDA GPU. Ollama requires driver 550+, or 570+ for compute capability 5.0–6.2; update the driver before relying on GPU acceleration.
This is a hypothetical PC. No inference benchmark was run.
GB uses 1,024³ bytes. Memory includes weights, context cache and runtime overhead.
Entire model runs from VRAM; no offload to system RAM required.
Storage: Pass
500.0 GB free, 60 MB required.
Acceleration: Pass
cuda acceleration available.
Assumptions and missing information
KV cache size is estimated using a fallback formula because architecture metadata is incomplete. Actual memory may differ significantly.
The NVIDIA driver version was not measured for every CUDA GPU. Ollama requires driver 550+, or 570+ for compute capability 5.0–6.2; update the driver before relying on GPU acceleration.
This is a hypothetical PC. No inference benchmark was run.
GB uses 1,024³ bytes. Memory includes weights, context cache and runtime overhead.
Exceeds documented context maximum (2,048). Memory estimate only.
Est. memory: 0.6 GB
Download: 0.1 GB
0.1 GB
0.6 GB
Excellent fit
low confidence
How this result was calculated
GPU memory: Pass
Model fully fits in usable GPU VRAM.
System RAM: Pass
Entire model runs from VRAM; no offload to system RAM required.
Storage: Pass
500.0 GB free, 101 MB required.
Acceleration: Pass
cuda acceleration available.
Assumptions and missing information
KV cache size is estimated using a fallback formula because architecture metadata is incomplete. Actual memory may differ significantly.
The NVIDIA driver version was not measured for every CUDA GPU. Ollama requires driver 550+, or 570+ for compute capability 5.0–6.2; update the driver before relying on GPU acceleration.
This is a hypothetical PC. No inference benchmark was run.
GB uses 1,024³ bytes. Memory includes weights, context cache and runtime overhead.
Entire model runs from VRAM; no offload to system RAM required.
Storage: Pass
500.0 GB free, 101 MB required.
Acceleration: Pass
cuda acceleration available.
Assumptions and missing information
KV cache size is estimated using a fallback formula because architecture metadata is incomplete. Actual memory may differ significantly.
The NVIDIA driver version was not measured for every CUDA GPU. Ollama requires driver 550+, or 570+ for compute capability 5.0–6.2; update the driver before relying on GPU acceleration.
This is a hypothetical PC. No inference benchmark was run.
GB uses 1,024³ bytes. Memory includes weights, context cache and runtime overhead.
Exceeds documented context maximum (2,048). Memory estimate only.
Est. memory: 0.8 GB
Download: 0.2 GB
0.2 GB
0.8 GB
Excellent fit
low confidence
How this result was calculated
GPU memory: Pass
Model fully fits in usable GPU VRAM.
System RAM: Pass
Entire model runs from VRAM; no offload to system RAM required.
Storage: Pass
500.0 GB free, 227 MB required.
Acceleration: Pass
cuda acceleration available.
Assumptions and missing information
KV cache size is estimated using a fallback formula because architecture metadata is incomplete. Actual memory may differ significantly.
The NVIDIA driver version was not measured for every CUDA GPU. Ollama requires driver 550+, or 570+ for compute capability 5.0–6.2; update the driver before relying on GPU acceleration.
This is a hypothetical PC. No inference benchmark was run.
GB uses 1,024³ bytes. Memory includes weights, context cache and runtime overhead.
Exceeds documented context maximum (2,048). Memory estimate only.
Est. memory: 0.9 GB
Download: 0.3 GB
0.3 GB
0.9 GB
Excellent fit
low confidence
How this result was calculated
GPU memory: Pass
Model fully fits in usable GPU VRAM.
System RAM: Pass
Entire model runs from VRAM; no offload to system RAM required.
Storage: Pass
500.0 GB free, 262 MB required.
Acceleration: Pass
cuda acceleration available.
Assumptions and missing information
KV cache size is estimated using a fallback formula because architecture metadata is incomplete. Actual memory may differ significantly.
The NVIDIA driver version was not measured for every CUDA GPU. Ollama requires driver 550+, or 570+ for compute capability 5.0–6.2; update the driver before relying on GPU acceleration.
This is a hypothetical PC. No inference benchmark was run.
GB uses 1,024³ bytes. Memory includes weights, context cache and runtime overhead.
Entire model runs from VRAM; no offload to system RAM required.
Storage: Pass
500.0 GB free, 287 MB required.
Acceleration: Pass
cuda acceleration available.
Assumptions and missing information
KV cache size is estimated using a fallback formula because architecture metadata is incomplete. Actual memory may differ significantly.
The NVIDIA driver version was not measured for every CUDA GPU. Ollama requires driver 550+, or 570+ for compute capability 5.0–6.2; update the driver before relying on GPU acceleration.
This is a hypothetical PC. No inference benchmark was run.
GB uses 1,024³ bytes. Memory includes weights, context cache and runtime overhead.
Entire model runs from VRAM; no offload to system RAM required.
Storage: Pass
500.0 GB free, 379 MB required.
Acceleration: Pass
cuda acceleration available.
Assumptions and missing information
KV cache size is estimated using a fallback formula because architecture metadata is incomplete. Actual memory may differ significantly.
The NVIDIA driver version was not measured for every CUDA GPU. Ollama requires driver 550+, or 570+ for compute capability 5.0–6.2; update the driver before relying on GPU acceleration.
This is a hypothetical PC. No inference benchmark was run.
GB uses 1,024³ bytes. Memory includes weights, context cache and runtime overhead.
Entire model runs from VRAM; no offload to system RAM required.
Storage: Pass
500.0 GB free, 379 MB required.
Acceleration: Pass
cuda acceleration available.
Assumptions and missing information
KV cache size is estimated using a fallback formula because architecture metadata is incomplete. Actual memory may differ significantly.
The NVIDIA driver version was not measured for every CUDA GPU. Ollama requires driver 550+, or 570+ for compute capability 5.0–6.2; update the driver before relying on GPU acceleration.
This is a hypothetical PC. No inference benchmark was run.
GB uses 1,024³ bytes. Memory includes weights, context cache and runtime overhead.
Entire model runs from VRAM; no offload to system RAM required.
Storage: Pass
500.0 GB free, 379 MB required.
Acceleration: Pass
cuda acceleration available.
Assumptions and missing information
KV cache size is estimated using a fallback formula because architecture metadata is incomplete. Actual memory may differ significantly.
The NVIDIA driver version was not measured for every CUDA GPU. Ollama requires driver 550+, or 570+ for compute capability 5.0–6.2; update the driver before relying on GPU acceleration.
This is a hypothetical PC. No inference benchmark was run.
GB uses 1,024³ bytes. Memory includes weights, context cache and runtime overhead.
Entire model runs from VRAM; no offload to system RAM required.
Storage: Pass
500.0 GB free, 379 MB required.
Acceleration: Pass
cuda acceleration available.
Assumptions and missing information
KV cache size is estimated using a fallback formula because architecture metadata is incomplete. Actual memory may differ significantly.
The NVIDIA driver version was not measured for every CUDA GPU. Ollama requires driver 550+, or 570+ for compute capability 5.0–6.2; update the driver before relying on GPU acceleration.
This is a hypothetical PC. No inference benchmark was run.
GB uses 1,024³ bytes. Memory includes weights, context cache and runtime overhead.
Entire model runs from VRAM; no offload to system RAM required.
Storage: Pass
500.0 GB free, 388 MB required.
Acceleration: Pass
cuda acceleration available.
Assumptions and missing information
KV cache size is estimated using a fallback formula because architecture metadata is incomplete. Actual memory may differ significantly.
The NVIDIA driver version was not measured for every CUDA GPU. Ollama requires driver 550+, or 570+ for compute capability 5.0–6.2; update the driver before relying on GPU acceleration.
This is a hypothetical PC. No inference benchmark was run.
GB uses 1,024³ bytes. Memory includes weights, context cache and runtime overhead.
Entire model runs from VRAM; no offload to system RAM required.
Storage: Pass
500.0 GB free, 498 MB required.
Acceleration: Pass
cuda acceleration available.
Assumptions and missing information
KV cache size is estimated using a fallback formula because architecture metadata is incomplete. Actual memory may differ significantly.
The NVIDIA driver version was not measured for every CUDA GPU. Ollama requires driver 550+, or 570+ for compute capability 5.0–6.2; update the driver before relying on GPU acceleration.
This is a hypothetical PC. No inference benchmark was run.
GB uses 1,024³ bytes. Memory includes weights, context cache and runtime overhead.
Entire model runs from VRAM; no offload to system RAM required.
Storage: Pass
500.0 GB free, 1.5 GB required.
Acceleration: Pass
cuda acceleration available.
Assumptions and missing information
KV cache size is estimated using a fallback formula because architecture metadata is incomplete. Actual memory may differ significantly.
The NVIDIA driver version was not measured for every CUDA GPU. Ollama requires driver 550+, or 570+ for compute capability 5.0–6.2; update the driver before relying on GPU acceleration.
This is a hypothetical PC. No inference benchmark was run.
GB uses 1,024³ bytes. Memory includes weights, context cache and runtime overhead.
Entire model runs from VRAM; no offload to system RAM required.
Storage: Pass
500.0 GB free, 537 MB required.
Acceleration: Pass
cuda acceleration available.
Assumptions and missing information
KV cache size is estimated using a fallback formula because architecture metadata is incomplete. Actual memory may differ significantly.
The NVIDIA driver version was not measured for every CUDA GPU. Ollama requires driver 550+, or 570+ for compute capability 5.0–6.2; update the driver before relying on GPU acceleration.
This is a hypothetical PC. No inference benchmark was run.
GB uses 1,024³ bytes. Memory includes weights, context cache and runtime overhead.
Entire model runs from VRAM; no offload to system RAM required.
Storage: Pass
500.0 GB free, 637 MB required.
Acceleration: Pass
cuda acceleration available.
Assumptions and missing information
KV cache size is estimated using a fallback formula because architecture metadata is incomplete. Actual memory may differ significantly.
The NVIDIA driver version was not measured for every CUDA GPU. Ollama requires driver 550+, or 570+ for compute capability 5.0–6.2; update the driver before relying on GPU acceleration.
This is a hypothetical PC. No inference benchmark was run.
GB uses 1,024³ bytes. Memory includes weights, context cache and runtime overhead.
Exceeds documented context maximum (2,048). Memory estimate only.
Est. memory: 1.4 GB
Download: 0.6 GB
0.6 GB
1.4 GB
Excellent fit
low confidence
How this result was calculated
GPU memory: Pass
Model fully fits in usable GPU VRAM.
System RAM: Pass
Entire model runs from VRAM; no offload to system RAM required.
Storage: Pass
500.0 GB free, 638 MB required.
Acceleration: Pass
cuda acceleration available.
Assumptions and missing information
KV cache size is estimated using a fallback formula because architecture metadata is incomplete. Actual memory may differ significantly.
The NVIDIA driver version was not measured for every CUDA GPU. Ollama requires driver 550+, or 570+ for compute capability 5.0–6.2; update the driver before relying on GPU acceleration.
This is a hypothetical PC. No inference benchmark was run.
GB uses 1,024³ bytes. Memory includes weights, context cache and runtime overhead.
Entire model runs from VRAM; no offload to system RAM required.
Storage: Pass
500.0 GB free, 639 MB required.
Acceleration: Pass
cuda acceleration available.
Assumptions and missing information
KV cache size is estimated using a fallback formula because architecture metadata is incomplete. Actual memory may differ significantly.
The NVIDIA driver version was not measured for every CUDA GPU. Ollama requires driver 550+, or 570+ for compute capability 5.0–6.2; update the driver before relying on GPU acceleration.
This is a hypothetical PC. No inference benchmark was run.
GB uses 1,024³ bytes. Memory includes weights, context cache and runtime overhead.
Entire model runs from VRAM; no offload to system RAM required.
Storage: Pass
500.0 GB free, 639 MB required.
Acceleration: Pass
cuda acceleration available.
Assumptions and missing information
KV cache size is estimated using a fallback formula because architecture metadata is incomplete. Actual memory may differ significantly.
The NVIDIA driver version was not measured for every CUDA GPU. Ollama requires driver 550+, or 570+ for compute capability 5.0–6.2; update the driver before relying on GPU acceleration.
This is a hypothetical PC. No inference benchmark was run.
GB uses 1,024³ bytes. Memory includes weights, context cache and runtime overhead.