Skip to main content

Documentation Index

Fetch the complete documentation index at: https://docs.axioniclabs.ai/llms.txt

Use this file to discover all available pages before exploring further.

What is Spectra?

Spectra is Axionic’s web workspace for training, adapting, and operating small language models. You can fine-tune a model for your domain, shape runtime behavior without redeploying, and monitor usage from the same product surface.

Core workflow

Train a model — Define tools or upload a text dataset, choose a teacher and student, and run supervised fine-tuning with an optional GRPO reinforcement phase. Optimize runtime behavior — Use the Optimization workspace to select a model, attach steering vectors, adjust decoding controls, run test generations, and produce API-ready request snippets. Add behavior safeguards — Create SAE-based behaviors that monitor for specific drift patterns and optionally pair them with corrective vectors. Operate from one workspace — Track model inventory, credits, billing, API keys, and runtime activity without leaving Spectra. Integrate anywhere — Every ready model can be called through the OpenAI-compatible API, the Mechanex SDK, or your own client code.

Access

Create an account at app.axioniclabs.ai with email/password or Google/GitHub OAuth.