Microsoft

About Microsoft

Explore research at Microsoft, a site featuring the impact of research along with publications, products, downloads, and research careers.

Microsoft Research conducts fundamental science and technology research across a spectrum of research areas. With labs around the globe we pursue breakthroughs across the computing and AI stack to accelerate discovery across sciences and advance the frontiers of computing and AI to create new foundations for others to build upon.

Artificial intelligence | Search & information retrieval | Human language technologies | Human-computer interaction | Computer vision | Graphics & multimedia | Audio & acoustics

Medical, health & genomics | Economics | Social sciences | Technology for emerging markets | Ecology & environment

Website ↗Documentation ↗

Source ↗ · 2026-11-12T02:17:04.301031227Z

669 indexed models

669 matching models

openai-whisper-large-v3-turbo-cuda-gpu · azureml v3

azureFoundry/azureml/openai-whisper-large-v3-turbo-cuda-gpu@3

Whisper Large V3 Turbo is an advanced speech recognition model, optimized for high-performance GPU inference. It is suitable for automatic speech recognition (ASR) tasks in various domains, leveraging large-scale training data for robust multilingual transcription. This model is

Maximum output
2,048
Input modalities
audio
Output modalities
text
Capabilities
audioInput

openai-whisper-large-v3-turbo-generic-cpu · azureml v3

azureFoundry/azureml/openai-whisper-large-v3-turbo-generic-cpu@3

Whisper Large V3 Turbo is an advanced speech recognition model, optimized for high-performance CPU inference. It is suitable for automatic speech recognition (ASR) tasks in various domains, leveraging large-scale training data for robust multilingual transcription. This model is

Maximum output
2,048
Input modalities
audio
Output modalities
text
Capabilities
audioInput

openai-whisper-large-v3-turbo-generic-cpu · azureml v4

azureFoundry/azureml/openai-whisper-large-v3-turbo-generic-cpu@4

Whisper Large V3 Turbo is an advanced speech recognition model, optimized for high-performance CPU inference. It is suitable for automatic speech recognition (ASR) tasks in various domains, leveraging large-scale training data for robust multilingual transcription. This model is

Maximum output
2,048
Input modalities
audio
Output modalities
text
Capabilities
audioInput

openai-whisper-medium-cuda-gpu · azureml v3

azureFoundry/azureml/openai-whisper-medium-cuda-gpu@3

This model is an optimized version of Whisper Medium for local inference. Optimized models are published here in ONNX format to run on CUDA devices, including server platforms, desktops, and mobile, with the precision best suited to each of these targets. > Review the [original

Maximum output
2,048
Input modalities
audio
Output modalities
text
Capabilities
audioInput

openai-whisper-medium-generic-cpu · azureml v3

azureFoundry/azureml/openai-whisper-medium-generic-cpu@3

This model is an optimized version of Whisper Medium for local inference. Optimized models are published here in ONNX format to run on CPU devices, including server platforms, desktops, and mobile, with the precision best suited to each of these targets. > Review the [original m

Maximum output
2,048
Input modalities
audio
Output modalities
text
Capabilities
audioInput

openai-whisper-medium-generic-cpu · azureml v4

azureFoundry/azureml/openai-whisper-medium-generic-cpu@4

This model is an optimized version of Whisper Medium for local inference. Optimized models are published here in ONNX format to run on CPU devices, including server platforms, desktops, and mobile, with the precision best suited to each of these targets. > Review the [original m

Maximum output
2,048
Input modalities
audio
Output modalities
text
Capabilities
audioInput

openai-whisper-small-cuda-gpu · azureml v3

azureFoundry/azureml/openai-whisper-small-cuda-gpu@3

This model is an optimized version of Whisper Small for local inference. Optimized models are published here in ONNX format to run on CUDA devices, including server platforms, desktops, and mobile, with the precision best suited to each of these targets. > Review the [original m

Maximum output
2,048
Input modalities
audio
Output modalities
text
Capabilities
audioInput

openai-whisper-small-generic-cpu · azureml v3

azureFoundry/azureml/openai-whisper-small-generic-cpu@3

This model is an optimized version of Whisper Small for local inference. Optimized models are published here in ONNX format to run on CPU devices, including server platforms, desktops, and mobile, with the precision best suited to each of these targets. > Review the [original mo

Maximum output
2,048
Input modalities
audio
Output modalities
text
Capabilities
audioInput

openai-whisper-small-generic-cpu · azureml v4

azureFoundry/azureml/openai-whisper-small-generic-cpu@4

This model is an optimized version of Whisper Small for local inference. Optimized models are published here in ONNX format to run on CPU devices, including server platforms, desktops, and mobile, with the precision best suited to each of these targets. > Review the [original mo

Maximum output
2,048
Input modalities
audio
Output modalities
text
Capabilities
audioInput

openai-whisper-tiny-cuda-gpu · azureml v3

azureFoundry/azureml/openai-whisper-tiny-cuda-gpu@3

This model is an optimized version of Whisper Tiny for local inference. Optimized models are published here in ONNX format to run on CUDA devices, including server platforms, desktops, and mobile, with the precision best suited to each of these targets. > Review the [original mo

Maximum output
2,048
Input modalities
audio
Output modalities
text
Capabilities
audioInput

openai-whisper-tiny-generic-cpu · azureml v4

azureFoundry/azureml/openai-whisper-tiny-generic-cpu@4

This model is an optimized version of Whisper Tiny for local inference. Optimized models are published here in ONNX format to run on CPU devices, including server platforms, desktops, and mobile, with the precision best suited to each of these targets. > Review the [original mod

Maximum output
2,048
Input modalities
audio
Output modalities
text
Capabilities
audioInput

parakeet-tdt-0.6b-v2-cuda-gpu · azureml v1

azureFoundry/azureml/parakeet-tdt-0.6b-v2-cuda-gpu@1

This model is an optimized version of parakeet-tdt-0.6b-v2 for local inference on CUDA GPUs. This optimized model is published here in ONNX format to run on CUDA-capable GPU devices, with the precision best suited to this target. # Model Description - **Developed by:** Microsoft

Maximum output
2,048
Input modalities
audio
Output modalities
text
Capabilities
audioInput

parakeet-tdt-0.6b-v2-generic-cpu · azureml v1

azureFoundry/azureml/parakeet-tdt-0.6b-v2-generic-cpu@1

This model is an optimized version of parakeet-tdt-0.6b-v2 for local inference. Optimized models are published here in ONNX format to run on CPU across devices, including server platforms, Windows, Linux and Mac desktops, and mobile CPUs. # Model Description - **Developed by:**

Maximum output
2,048
Input modalities
audio
Output modalities
text
Capabilities
audioInput

Paza MMS-1B-All · azureml-paza-p v1

azureFoundry/azureml-paza-p/paza-mms-1b-all@1

Fine-tuned MMS-1B-all (Wav2Vec2 + language adapters) for automatic speech recognition in six Kenyan languages. Non-commercial (CC-BY-NC-4.0).

Input modalities
audio
Output modalities
text
Capabilities
audioInput

Paza Phi-4 Multimodal Instruct · azureml-msr v1

azureFoundry/azureml-msr/paza-Phi-4-multimodal-instruct@1

Fine-tuned Phi-4-multimodal-instruct for automatic speech recognition in six Kenyan languages (Swahili, Kalenjin, Kikuyu, Luo, Maasai, Somali); retains the base model's multimodal capabilities.

Input modalities
audio, image, text
Output modalities
text
Capabilities
audioInput, multimodal, visionInput

Phi-3.5-mini-instruct

vertexAI/publishers/microsoft/models/phi3@phi-3.5-mini-instruct

Input modalities
text
Output modalities
text

Benchmarks

  • Open LLM Leaderboard v2: 28.184%

Models

Loading
EVIDENCE / COMPARISON

Benchmarks

Explore every benchmark

Loading

These charts contain model-linked results. The charts above also include unlinked source profiles. Compare scores within each evaluation. Model quality, usage, transparency and system performance measure different things.

Install TensorTrace