- TRAINING
- INVENTORY
- CONTROL
- HUMAN
Short answer
How can I fine tune custom models with my AI?
You can trigger training jobs and manage specialized model instances by giving your agent access to Gradient AI. This allows you to use your own datasets to create custom models through simple text commands. You'll get a breakdown of the training and management workflow below.
Model management outcomes
What you can achieve
These are the results of your requests.
TRAINING
New custom models
Your agent starts a training job using your specific datasets. You get a specialized model instance ready for use.
INVENTORY
Model oversight
The AI checks your current list of models. You can see exactly what you have active in your account.
CONTROL
Instance management
You can pull specific details for a model or remove old ones. This keeps your model library clean and organized.
HUMAN
Dataset preparation
The AI cannot clean or format your raw data files. You must prepare your training datasets before asking the agent to start a job.
The workflow
What your AI does when you request training.
The agent handles the heavy lifting of interacting with the Gradient AI API.
Start training
The agent initiates a training process using the datasets you provide.
fine_tune_modelCreate instances
The agent sets up new model instances based on your training results.
create_modelCheck status
The agent looks up your existing models to check their details or availability.
list_modelsget_modelCleanup
The agent removes specific model instances you no longer need.
delete_model
Try it
Copy these to start.
Use these prompts to kick off your first training job.
Starting points
Replace the bracketed text with your actual dataset names or model IDs.
Gradient AI Connector
You're all set. Choose your MCP client and follow the setup instructions.
https://edge.vinkius.com/vk_preview_KwJfHerXBfsWKXVQH8bGJG4HGUwUfVlRXkPpGPNU/mcpClaude Desktop
Follow the steps below to connect in seconds.
- 1In Claude Desktop, open Settings → Connectors.
- 2Click “Add custom connector” and paste the connector link above as the remote MCP server URL.
- 3Click Add and start a new chat — Gradient AI capabilities are ready to use.
{
"mcpServers": {
"gradient-ai-llm-api-finetuning-mcp": {
"url": "https://edge.vinkius.com/vk_preview_KwJfHerXBfsWKXVQH8bGJG4HGUwUfVlRXkPpGPNU/mcp"
}
}
}Fine tune a new model using my dataset named [dataset_name].
List all my current models in Gradient AI.
Show me the details for the model with ID [model_id].
Delete the model named [model_name] once the training is done.
Claude
ChatGPT
Cursor
VS Code
Windsurf
Claude Code
JetBrains
Cline
Start here
Connect Gradient AI once, then ask.
One connection links your agent to the Gradient AI API. Your credentials stay encrypted and secure on our end.
Connect Gradient AI to your AIFAQ
How this task behaves
- 01
Can the AI format my JSON training data?
No. You must provide correctly formatted datasets; the agent only triggers the training job.
- 02
Will the AI delete my models without asking?
No. The agent only deletes a model if you explicitly command it to do so.
- 03
Can I monitor training progress in real time?
The agent can check the status of models, but it does not provide a live streaming progress bar.
- 04
How do I see which models I have already trained?
Just ask your agent to list your models.
- 05
Can the AI create a model without a training job?
Yes, the agent can create model instances directly if the tool allows.
More questions about Gradient AI? The Connector page answers them. See everything the Gradient AI Connector can do
Connect Gradient AI to Claude, Cursor, ChatGPT & more
