Loading page content
Local Models
Read browser-visible hardware information, estimate local inference memory requirements, and prioritize the highest-scoring open-weight models this device can run.
Recommended models
Current Device
Only browser-visible hardware information is used; all detection and estimates stay on this page.
- CPU threads
- Not exposed
- Memory
- Not exposed
- VRAM
- Not inferred
- Graphics API
- WebGPU unavailable
- Browser storage quota
- Not exposed
GPUSpecific model not exposed
PlatformNot exposed
Runtime Budget
Estimate local runtime requirements from quantized weights, inference overhead, and system reserve.
Quantization
Q4 is the default because it is the most common capacity baseline for local large-model deployment.
Available memory
No model recommendation yet
The browser did not expose memory information and VRAM could not be inferred from the GPU model. Select available memory manually to see recommendations.
Candidate Models
Shows open-weight models that have both an overall score and a recorded parameter count, ordered by score.
| Model | Score | Parameters | Memory estimate | Status |
|---|---|---|---|---|
| 100.0 | 2800B / 104B active | 1527 GB | Pending detection | |
| 94.8 | 744B / 40B active | 416.8 GB | Pending detection | |
| 94.5 | 2400B / 95B active | 1311 GB | Pending detection | |
| 93.0 | 1650.5B / 49B active | 906.3 GB | Pending detection | |
| 92.5 | 320B / 18B active | 187.8 GB | Pending detection | |
| 90.9 | 125B / 6B active | 81.8 GB | Pending detection | |
| 90.4 | 744B / 40B active | 416.8 GB | Pending detection | |
| 87.9 | 1000B / 32B active | 555 GB | Pending detection | |
| 87.4 | 284B / 13B active | 168.4 GB | Pending detection | |
| 86.1 | 27.8B | 20.6 GB | Pending detection | |
| 86.1 | 1000B / 32B active | 555 GB | Pending detection | |
| 85.2 | 295B / 21B active | 174.3 GB | Pending detection |
