Phi-4-reasoning-vision-15B
Phi-4-reasoning-vision-15B is Microsoft's March 2026 multimodal Phi model, which adds a SigLIP-2 vision encoder to Phi-4-reasoning. A single model switches between step-by-step reasoning for math and science and direct answers for perception tasks such as OCR and GUI grounding. It is MIT licensed and focused on English.
Specifications
- Context window
- 16Kabout 12,300 words, or 25 pages
- Max output
- Not published
- Input price
- —no list price confirmed
- Output price
- —
- Cached input
- Not published
- Weights
- OpenMIT
- Parameters
- 15.1Bdense
- Input types
- Text, Images
- Output types
- Text
- Reasoning mode
- Yes
- Tool calling
- Not documented
- Released
- Mar 4, 2026Current
No serverless price found on the official pricing pages of Together AI, Fireworks, DeepInfra or Groq as of 2026-10-10.
What it costs in practice
| Workload | Monthly cost at list price |
|---|---|
| Support chatbot30,000 conversations a month, 2,000 input and 400 output tokens each | No list price |
| Document summaries2,000 documents a month, 12,000 input and 800 output tokens each | No list price |
| Coding agent1,000 tasks a month, 60,000 input and 6,000 output tokens each | No list price |
Standard rates, no caching or batch discounts. Run your own numbers.
Strengths
- MIT license
- Reasoning and direct-answer modes in one model
- Trained for screen understanding and GUI element grounding
Trade-offs to weigh
- 16K context window limits long documents
- Primarily English; tool calling is not documented
Where it fits
Microsoft (Phi) positions Phi-4-reasoning-vision-15B for:
Where you can use it
- Hugging Face
Run it yourself
Released under MIT. Approximate memory needed just to load the weights:
| Precision | Memory for weights |
|---|---|
| 16-bit (BF16/FP16) | 30 GB |
| 8-bit (FP8/INT8) | 15 GB |
| 4-bit (Q4_K_M, about 4.85 bits) | 9 GB |
Add memory for the KV cache (it grows with context length and concurrent users) and 1-2 GB of runtime overhead. Size it precisely in the VRAM calculator.
Alternatives to consider
| Model | Tier | Context | Input / output per 1M | |
|---|---|---|---|---|
| Ministral 3 8BMistral AI | Compact | 256K | $0.15 / $0.15 | Compare |
| Ministral 3 3BMistral AI | Compact | 256K | $0.10 / $0.10 | Compare |
| Qwen3.6-35B-A3BAlibaba (Qwen) | Compact | 256K | $0.10 / $0.95 | Compare |
| Gemma 4 26B A4BGoogle | Compact | 256K | $0.07 / $0.34 | Compare |
Frequently asked questions
How much does Phi-4-reasoning-vision-15B cost?
We couldn't confirm a published per-token list price for Phi-4-reasoning-vision-15B on October 10, 2026. No serverless price found on the official pricing pages of Together AI, Fireworks, DeepInfra or Groq as of 2026-10-10. Check the provider's pricing page for current rates.
What is the context window of Phi-4-reasoning-vision-15B?
Phi-4-reasoning-vision-15B accepts up to 16,384 tokens of context (about 12,300 words, or 25 pages). Long prompts cost more and can be slower, so most apps send only what the task needs.
Is Phi-4-reasoning-vision-15B open source?
Phi-4-reasoning-vision-15B is an open-weight model released under the MIT license. You can download the weights from Hugging Face (microsoft/Phi-4-reasoning-vision-15B) and run or fine-tune them yourself, within the license terms. Open weights isn't the same as fully open source: training data and code may not be published.
Can I run Phi-4-reasoning-vision-15B locally?
Yes, if you have enough memory. The weights alone need about 9 GB at 4-bit quantization and about 30 GB at 16-bit, plus room for the KV cache. Use the VRAM calculator on LLModel.com to check a specific GPU or Mac.
What is Phi-4-reasoning-vision-15B best for?
Microsoft (Phi) positions Phi-4-reasoning-vision-15B for multimodal, reasoning, self-hosting, research. It accepts text, images as input, has a reasoning (thinking) mode and doesn't document tool calling. Test it on a sample of your own tasks before you choose it for production.
Sources
- huggingface.co/microsoft/Phi-4-reasoning-vision-15B
- huggingface.co/microsoft/Phi-4-reasoning-vision-15B/raw/main/config.json
- huggingface.co/api/models/microsoft/Phi-4-reasoning-vision-15B
Facts checked against these pages on October 10, 2026. Spot an error? Tell us.
Still choosing a model?
Send us your use case and constraints. We'll come back with a shortlist and an evaluation plan you can run on your own data.
- Free and vendor-neutral: no model provider pays for placement.
- Your details go only to partners you agree to.