Description
d3-turbo is the latest iteration of the Decision 3.0 models from vLLM Semantic Router, designed to enhance decision-making processes through its 27 billion parameter multimodal capabilities. This model can handle inputs in the form of text, JSON, images, and videos, allowing users to pose various types of questions such as choice, yes/no, score, and location-based queries in images or videos. It processes each input once per request, providing answers with associated probabilities, which makes it highly efficient for handling multiple questions simultaneously.
Read More
The model supports a long context of up to 131,072 tokens per question when configured with long_context=True, with a default context of 32,768 tokens. This feature is particularly useful for complex queries that require extensive input data. d3-turbo is optimized for performance, outperforming its predecessor, d3-flash, in scenarios involving multiple questions about the same input, making it significantly faster and more efficient.
d3-turbo is available under the Apache-2.0 license and can be accessed via a free API, which allows users to test its capabilities without any initial cost. The API supports up to 64 questions per request, with input limits of 131,072 characters and 16 images or one video per request. This model is particularly beneficial for industries requiring precise decision-making capabilities, such as media analysis, content moderation, and automated reporting.
The model's performance is evidenced by its scores on the Jev Decision Index, where it surpasses previous models in both text and vision tasks. It achieves a text score of 70.2 and a vision score of 72.9, indicating its robust capability in handling diverse decision-making tasks. With its advanced features and open-source availability, d3-turbo is positioned as a powerful tool for developers and organizations looking to integrate sophisticated decision-making models into their workflows.
- CategoryAI Models & LLMs
- Pricing ModelFreemium
- DeploymentCloud
- PlatformsWeb
d3-turbo Highlights
27B multimodal decision model
Processes text, JSON, images, and videos
Answers with probabilities
Long context support up to 131,072 tokens
Apache-2.0 license
Free API access
Efficient performance for multiple questions
High scores on Jev Decision Index
Getting Started with d3-turbo
Access model: Visit decision.vllm-sr.ai
Authenticate: Obtain a free API key
Set up environment: Configure API settings
Integrate via API: Use API to send requests
Optimise: Adjust parameters for performance
d3-turbo's Use Cases
- Media Analysis
- Content Moderation
- Automated Reporting
- Image Recognition
- Video Tracking
