Phi-4-multimodal-instruct is an open-source multimodal model from Microsoft that processes text, images, and audio in a unified architectural solution. It builds upon the Phi-3.5 and Phi-4.0 technologies, featuring an extended context window of 128K tokens and support for 23 languages in text (including Russian), 8 languages in audio, and English for visual tasks. The model is optimized for environments with limited computational resources and low-latency scenarios, demonstrating strong performance in mathematics, logic, speech recognition, translation, and image analysis.
A single neural network handles text, images (OCR, tables, diagrams), and audio (recognition, translation, summarization). For example, in DocVQA benchmarks, the model achieves 93.2% accuracy, outperforming Gemini-2.0-Flash (92.1%).
The model is ideal for multisensory applications—joint processing of audio and images (e.g., video analysis with subtitles). At the same time, thanks to optimization via Microsoft Olive and ONNX GenAI Runtime, it can be deployed on edge devices, including smartphones and IoT systems, even with limited computational resources.
| Model Name | Context | Type | GPU | Status | Link |
|---|
There are no public endpoints for this model yet.
Rent your own physically dedicated instance with hourly or long-term monthly billing.
We recommend deploying private instances in the following scenarios:
| Name | GPU | TPS | Max Concurrency | |||
|---|---|---|---|---|---|---|
131,072.0 |
1 | $0.53 | 1.032 | Launch | ||
131,072.0 tensor |
2 | $0.54 | 1.325 | Launch | ||
131,072.0 tensor |
2 | $0.57 | 1.325 | Launch | ||
131,072.0 |
1 | $0.83 | 1.032 | Launch | ||
131,072.0 tensor |
3 | $0.84 | 1.225 | Launch | ||
131,072.0 |
1 | $1.02 | 1.032 | Launch | ||
131,072.0 |
1 | $1.20 | 1.482 | Launch | ||
131,072.0 tensor |
2 | $1.23 | 2.225 | Launch | ||
131,072.0 tensor |
3 | $1.43 | 1.057 | Launch | ||
131,072.0 |
1 | $1.59 | 1.482 | Launch | ||
131,072.0 |
1 | $2.37 | 4.182 | Launch | ||
131,072.0 |
1 | $3.83 | 4.182 | Launch | ||
131,072.0 |
1 | $4.11 | 4.969 | Launch | ||
131,072.0 |
1 | $4.74 | 7.613 | Launch | ||
| Name | GPU | TPS | Max Concurrency | |||
|---|---|---|---|---|---|---|
131,072.0 tensor |
2 | $0.54 | 1.163 | Launch | ||
131,072.0 tensor |
2 | $0.57 | 1.163 | Launch | ||
131,072.0 tensor |
3 | $0.84 | 1.063 | Launch | ||
131,072.0 tensor |
2 | $0.93 | 2.063 | Launch | ||
131,072.0 |
1 | $1.20 | 1.320 | Launch | ||
131,072.0 tensor |
2 | $1.23 | 2.063 | Launch | ||
131,072.0 tensor |
2 | $1.56 | 2.063 | Launch | ||
131,072.0 |
1 | $1.59 | 1.320 | Launch | ||
131,072.0 tensor |
4 | $1.82 | 1.301 | Launch | ||
131,072.0 tensor |
2 | $1.92 | 2.063 | Launch | ||
131,072.0 |
1 | $2.37 | 4.020 | Launch | ||
131,072.0 |
1 | $3.83 | 4.020 | Launch | ||
131,072.0 |
1 | $4.11 | 4.807 | Launch | ||
131,072.0 |
1 | $4.74 | 7.451 | Launch | ||
| Name | GPU | TPS | Max Concurrency | |||
|---|---|---|---|---|---|---|
131,072.0 tensor |
3 | $0.88 | 1.534 | Launch | ||
131,072.0 tensor |
2 | $0.93 | 1.691 | Launch | ||
131,072.0 tensor |
3 | $1.06 | 1.534 | Launch | ||
131,072.0 tensor |
4 | $1.12 | 1.153 | Launch | ||
131,072.0 tensor |
2 | $1.23 | 1.691 | Launch | ||
131,072.0 tensor |
2 | $1.56 | 1.691 | Launch | ||
131,072.0 tensor |
2 | $1.92 | 1.691 | Launch | ||
131,072.0 tensor |
2 | $2.22 | 2.591 | Launch | ||
131,072.0 |
1 | $2.37 | 3.647 | Launch | ||
131,072.0 tensor |
2 | $2.93 | 2.591 | Launch | ||
131,072.0 |
1 | $3.83 | 3.647 | Launch | ||
131,072.0 |
1 | $4.11 | 4.434 | Launch | ||
131,072.0 |
1 | $4.74 | 7.078 | Launch | ||
Contact our dedicated neural networks support team at nn@immers.cloud or send your request to the sales department at sale@immers.cloud.