Cohere
Label Studio Enterprise integrates with Cohere through the Prompts interface so you can call models like command-r and command-r-plus directly from your labeling projects. Once you add Cohere as a provider in Prompts, its models appear alongside OpenAI and Gemini, letting you generate annotations, test prompts, and evaluate outputs against ground truth inside Label Studio.
With this connection, you can:
- Pre-label tasks using Cohere models and write results into project controls (Choices, Labels, TextArea, etc.)
- Iterate on prompts using real project data before scaling to bulk auto-labeling
- Compare model behavior and quality across providers while keeping your workflow in one place
This Cohere integration is available in Label Studio Enterprise via Prompts.
How Label Studio Connects with Cohere in Prompts
Label Studio Enterprise treats each Prompt as a reusable configuration that specifies which provider to use, which model to call, and how to map inputs and outputs. When you select Cohere as the provider and run a Prompt:
- The Prompts UI sends your selected tasks, prompt template, and chosen Cohere model to the Enterprise backend
- The backend calls Cohere with your API key and task data
- Cohere’s response is written back into your project as predictions or displayed in the Prompt response pane
- You can review outputs, refine your prompt, and then apply it at scale to pre-annotate tasks
