ClaudeVision<T>

Models & Types · AiDotNet.VisionLanguage.Proprietary

Claude Vision: reference implementation of Anthropic’s multimodal reasoning model.

Inheritance: NeuralNetworkBase<T> → VisionLanguageModelBase<T> → ClaudeVision<T>

Implements: IProprietaryVLM<T>

For Beginners

Claude Vision is a proprietary multimodal model from Anthropic with strong visual reasoning and document understanding. Default values follow the model’s recommended settings.

How It Works

Claude Vision is a reference implementation of Anthropic’s multimodal reasoning model. The Claude 3/4 family features strong document and chart understanding, extended thinking for complex visual reasoning, and processes images alongside text in a unified transformer architecture with safety-focused RLHF training.

References:

  • Claude 3/4 Vision: strong document and chart understanding with extended thinking (Anthropic, 2024-2025)

Example

// Create a Claude Vision model for multimodal reasoning
// with strong document and chart understanding capabilities
var architecture = new NeuralNetworkArchitecture<double>(
    inputType: InputType.TwoDimensional,
    taskType: NeuralNetworkTaskType.ImageClassification,
    inputHeight: 224, inputWidth: 224, inputDepth: 3, outputSize: 512);
            
// ONNX inference mode with pre-trained model
var model = new ClaudeVision<double>(architecture, "claudevision.onnx");
            
// Training mode with native layers
var trainModel = new ClaudeVision<double>(architecture, new ClaudeVisionOptions());

Constructors

Constructor Summary
ClaudeVision(NeuralNetworkArchitecture<T> architecture, ClaudeVisionOptions options, IGradientBasedOptimizer<T, Tensor<T>, Tensor<T>> optimizer)
ClaudeVision(NeuralNetworkArchitecture<T> architecture, string modelPath, ClaudeVisionOptions options)

Properties

Property Summary
int DecoderEmbeddingDim { get; }
int EmbeddingDimension { get; }
int MaxGenerationLength { get; }
string LanguageModelName { get; }
string Provider { get; }

Methods

Method Summary
ModelMetadata<T> GetModelMetadata()
Tensor<T> Chat(Tensor<T> image, string prompt)
Tensor<T> EncodeImage(Tensor<T> image)
Tensor<T> GenerateFromImage(Tensor<T> image, string prompt) Generates text from an image using Claude’s multimodal reasoning architecture.