[CLI] foundry model ls only showing CPU models

[MorrisGallego]

Problem

When executing the foundry model ls command, only CPU models are shown, no NPU or GPU models available. Checking the log it seems that something bad is happening in the EP registration process. Looks like same error as in #797

Diagnostics

$> foundry server start

■ Starting Foundry Local server

CUDAExecutionProvider          ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100%
WebGpuExecutionProvider        ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━   0% ⠋
WebGpuExecutionProvider        ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100%
OpenVINOExecutionProvider      ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100%
NvTensorRTRTXExecutionProvider ━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━ 100%

● success: Server ready (http://127.0.0.1:51454)

$> foundry model ls
╭───────────────────────────────────┬────────────┬─────────┬────────┬───────┬────────╮
│ Model Name                        │ Type       │ Size    │ Device │ Tools │ Cached │
├───────────────────────────────────┼────────────┼─────────┼────────┼───────┼────────┤
│ deepseek-r1-14b                   │ Chat       │ 11.5 GB │ CPU    │ ○     │ ○      │
│ deepseek-r1-7b                    │ Chat       │ 6.4 GB  │ CPU    │ ○     │ ○      │
│ gpt-oss-20b                       │ Chat       │ 12.3 GB │ CPU    │ ○     │ ○      │
│ ministral-3-3b-instruct-2512      │ Multimodal │ 4.6 GB  │ CPU    │ ●     │ ○      │
│ mistral-7b-v0.2                   │ Chat       │ 4.1 GB  │ CPU    │ ○     │ ○      │
│ mistral-nemo-12b-instruct         │ Chat       │ 6.8 GB  │ CPU    │ ●     │ ○      │
│ nemotron-3.5-asr-streaming-0.6b   │ Speech     │ 756 MB  │ CPU    │ ○     │ ○      │
│ nemotron-speech-streaming-en-0.6b │ Speech     │ 696 MB  │ CPU    │ ○     │ ○      │
│ nemotron-speech-streaming-es-0.6b │ Speech     │ 729 MB  │ CPU    │ ○     │ ○      │
│ olmo-3-7b-instruct                │ Chat       │ 5.2 GB  │ CPU    │ ●     │ ○      │
│ phi-3-mini-128k                   │ Chat       │ 2.5 GB  │ CPU    │ ○     │ ○      │
│ phi-3-mini-4k                     │ Chat       │ 2.5 GB  │ CPU    │ ○     │ ○      │
│ phi-3.5-mini                      │ Chat       │ 2.5 GB  │ CPU    │ ○     │ ○      │
│ phi-4                             │ Chat       │ 10.2 GB │ CPU    │ ○     │ ○      │
│ phi-4-mini                        │ Chat       │ 4.8 GB  │ CPU    │ ●     │ ○      │
│ phi-4-mini-reasoning              │ Chat       │ 4.5 GB  │ CPU    │ ○     │ ○      │
│ phi-4-reasoning                   │ Chat       │ 10.2 GB │ CPU    │ ○     │ ○      │
│ qwen2.5-0.5b                      │ Chat       │ 822 MB  │ CPU    │ ●     │ ○      │
│ qwen2.5-1.5b                      │ Chat       │ 1.8 GB  │ CPU    │ ●     │ ○      │
│ qwen2.5-14b                       │ Chat       │ 11.1 GB │ CPU    │ ●     │ ○      │
│ qwen2.5-7b                        │ Chat       │ 6.2 GB  │ CPU    │ ●     │ ○      │
│ qwen2.5-coder-0.5b                │ Chat       │ 822 MB  │ CPU    │ ●     │ ○      │
│ qwen2.5-coder-1.5b                │ Chat       │ 1.8 GB  │ CPU    │ ●     │ ○      │
│ qwen2.5-coder-14b                 │ Chat       │ 11.1 GB │ CPU    │ ●     │ ○      │
│ qwen2.5-coder-7b                  │ Chat       │ 6.2 GB  │ CPU    │ ●     │ ○      │
│ qwen3-0.6b                        │ Chat       │ 593 MB  │ CPU    │ ●     │ ○      │
│ qwen3-1.7b                        │ Chat       │ 1.3 GB  │ CPU    │ ●     │ ○      │
│ qwen3-14b                         │ Chat       │ 9.3 GB  │ CPU    │ ●     │ ○      │
│ qwen3-4b                          │ Chat       │ 2.7 GB  │ CPU    │ ●     │ ○      │
│ qwen3-8b                          │ Chat       │ 5.7 GB  │ CPU    │ ●     │ ○      │
│ qwen3-embedding-0.6b              │ Embedding  │ 495 MB  │ CPU    │ ○     │ ○      │
│ qwen3-embedding-8b                │ Embedding  │ 5.6 GB  │ CPU    │ ○     │ ○      │
│ qwen3-vl-2b-instruct              │ Multimodal │ 1.3 GB  │ CPU    │ ●     │ ○      │
│ qwen3-vl-4b-instruct              │ Multimodal │ 2.7 GB  │ CPU    │ ●     │ ○      │
│ qwen3-vl-8b-instruct              │ Multimodal │ 5.0 GB  │ CPU    │ ●     │ ○      │
│ qwen3.5-0.8b                      │ Multimodal │ 904 MB  │ CPU    │ ●     │ ○      │
│ qwen3.5-2b                        │ Multimodal │ 2.1 GB  │ CPU    │ ●     │ ○      │
│ qwen3.5-2b-text                   │ Chat       │ 1.4 GB  │ CPU    │ ●     │ ○      │
│ qwen3.5-4b                        │ Multimodal │ 4.3 GB  │ CPU    │ ●     │ ○      │
│ qwen3.5-9b                        │ Multimodal │ 7.9 GB  │ CPU    │ ●     │ ○      │
│ smollm3-3b                        │ Chat       │ 2.1 GB  │ CPU    │ ●     │ ○      │
│ whisper-base                      │ Speech     │ 205 MB  │ CPU    │ ○     │ ○      │
│ whisper-large-v3-turbo            │ Speech     │ 1.3 GB  │ CPU    │ ○     │ ○      │
│ whisper-medium                    │ Speech     │ 937 MB  │ CPU    │ ○     │ ○      │
│ whisper-small                     │ Speech     │ 438 MB  │ CPU    │ ○     │ ○      │
│ whisper-tiny                      │ Speech     │ 131 MB  │ CPU    │ ○     │ ○      │
╰───────────────────────────────────┴────────────┴─────────┴────────┴───────┴────────╯
Type: Chat / Multimodal / Speech / Embedding = model task.
Device: NPU / GPU / CPU = default hardware target selected for this machine.
Tools: supports tool calling; the model can request API calls or app actions instead of only returning text.
Cached: model files are already downloaded on this machine.
Use --variants for every hardware-specific variant, provider, size, and cache state.

foundry.core.log

System

• OS: Microsoft Windows 10.0.26200
• Architecture: X64
• CPU: Intel(R) Core(TM) Ultra 7 265K (20 logical cores)
• GPU: Intel Corporation Intel(R) Graphics (2.0 GB), NVIDIA NVIDIA GeForce RTX 4070 SUPER (4.0 GB)
• NPU: Intel Corporation Intel(R) AI Boost
• RAM: 18.5 GB available / 31.4 GB total
• Disk: 773.0 GB free / 1.8 TB total (C:)
• .NET: .NET 9.0.17

Service

• State: Ready
• Web URLs: http://127.0.0.1:60155
• CLI version: 0.10.1

Models

• Available variants: 46
• Loaded: 0
• Cached: 0

Connectivity

• Local service: Reachable
• Model registry: Reachable

Warnings

• No warnings.

[samuel100]

Thanks for filing an issue @MorrisGallego! Indeed, this is the same issue as #797 - it looks like WebGpu is being picked up by both the Windows EP Catalog and the EP we pull. We have a fix and it will come in the next release.

Posting here, the details: #797 (comment)

[andrewtheart]

This is a pretty serious flaw - can I rely on Foundry to actually be embedded in my application?