Phi 3.5 Vision Instruct API
About
Integrate multimodal vision capabilities using Phi 3.5 Vision Instruct, designed to process both text and images seamlessly. This model delivers available completely free of charge, native tool calling support, open weights architecture. Access Phi 3.5 Vision Instruct via the Nvidia API with up to 4K output tokens.
Capabilities
Input Modalities
textimage
Output Modalities
text
Reasoning
No
Structured Output
Yes
Tool Use
Yes
WeightsOpen
Temperature
Adjustable
Attachment
Not Supported
Limits
Context Window
128K
Tokens
Input Limit
128K
Tokens
Max Output
4K
Tokens
Pricing
Standard (per 1M tokens)
Input
Free
Output
Free
Frequently Asked Questions
Phi 3.5 Vision Instruct by Nvidia costs Free per 1M input tokens and Free per 1M output tokens.
Model Details
API Integration
NPM Package
@ai-sdk/openai-compatible
Environment Variables
NVIDIA_API_KEY
API Base URL
https://integrate.api.nvidia.com/v1