MCP server for Fireworks AI — models, deployments, datasets and fine-tuning jobs. This server exposes model and experiment management endpoints, enabling orchestration of AI assets and training workflows in a centralized way.

🛠️ Key Features

Models, deployments, datasets, and fine-tuning job management
Centralized MCP protocol implementation for consistent interactions
Readme excerpt-based documentation for quick orientation
Slug: io-usefulapi-fireworks

🚀 Use Cases

Cataloging AI models and their deployment statuses
Managing datasets and fine-tuning experiments
Coordinating model lifecycle from training to production

⚡ Developer Benefits

Clear MCP server surface for integration with tooling
Standardized model deployment and experiment workflows
Lightweight, descriptive readme excerpt guidance

⚠️ Limitations

Details are derived from a concise description; may require consult of repository for full API surface
Limited field coverage in source data (name, description, slug, readmeExcerpt)

Fireworks AI MCP by usefulapi

Inspect Fireworks AI models, deployments, datasets and fine-tuning jobs from Claude, Cursor, or any MCP client. Hosted, no local install — connect with your Fireworks API key.

Live endpoint: https://fireworks.usefulapi.io/mcp · Homepage: https://usefulapi.io/fireworks

Add to Claude

json

{
  "mcpServers": {
    "fireworks": {
      "url": "https://fireworks.usefulapi.io/mcp"
    }
  }
}

On first connect you'll paste your Fireworks API key. It's validated, stored per-user, and scoped to you.

Tools

Tool	Type	What it does
`fireworks_get_account`	read	Get account info
`fireworks_list_models`	read	List models
`fireworks_get_model`	read	Get a model
`fireworks_list_deployments`	read	List deployments
`fireworks_get_deployment`	read	Get a deployment
`fireworks_list_deployed_models`	read	List deployed models
`fireworks_list_datasets`	read	List datasets
`fireworks_get_dataset`	read	Get a dataset
`fireworks_list_fine_tuning_jobs`	read	List fine-tuning jobs
`fireworks_get_fine_tuning_job`	read	Get a fine-tuning job
`fireworks_list_batch_inference_jobs`	read	List batch inference jobs
`fireworks_list_users`	read	List account users
`fireworks_create_dataset`	write	Create a dataset
`fireworks_delete_deployment`	write	Delete a deployment
`fireworks_usage_status`	meta	Your tier + usage this month
`fireworks_upgrade`	meta	Upgrade to Pro

Read tools are annotated readOnlyHint; write tools are destructiveHint so clients confirm before mutating.

Pricing

Plan	Price	Limit
Free	$0	100 tool calls / month
Pro	$9/mo or $90/yr (2 months free)	Unlimited

License

MIT — see LICENSE. Documentation only; the server is hosted.

Fireworks