Introduction
LLMTune is a platform for inference and fine-tuning of language models. You get a single control plane: manage API keys, run inference, launch fine-tuning jobs, and track usage and billing in one place.Core capabilities
- Inference — OpenAI-compatible chat and completion endpoints. Use your API key to call supported models with streaming and standard parameters.
- Agent — A coding agent with tools (e.g. read file, write file, run terminal). Use it from compatible IDEs or any OpenAI-compatible client; the API returns tool calls for the client to execute.
- Fine-tuning — Train on your data using platform-supported base models. Submit jobs via the API or the dashboard, monitor progress, and deploy trained models for inference.
- Billing and usage — Pay-as-you-go balance. Usage is metered by tokens; deductions happen per request. Low balance returns
402 Payment Required.
High-level architecture
- API — All programmatic access goes through the REST API. Authentication is via Bearer API key.
- Auth — API keys are created in the dashboard and grant access to your account (inference, training, usage).
- Balance — Credits are deducted per inference and training; top up via the dashboard.
- Training — Fine-tuning jobs run on platform infrastructure; you provide the dataset and configuration.
How to use this documentation
- Start here — Getting started for account setup, creating an API key, and your first request.
- Inference — Inference overview and Chat completions for calling models.
- Agent — Agent overview for the coding agent and tool use.
- Fine-tuning — Fine-tuning overview for datasets, workflow, and deployment.
- Billing — Billing & usage for how usage and deductions work.
- Errors — Errors & status codes for handling API errors.