Skip to main content

Documentation Index

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

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

Bitrecs V2 is a prompt evolution subnet on Bittensor that rewards miners who craft the most effective LLM prompts for ecommerce recommendation tasks. Miners submit artifact.yaml files — containing prompts, models, and sampling parameters — which are evaluated against a rotating set of real-world ecommerce scenarios. The top-performing artifact earns Bittensor emissions through a winner-takes-all scoring engine.

Start mining

Submit your first artifact and begin earning emissions on the Bitrecs subnet.

Run a validator

Set up a validator node to evaluate miner submissions and participate in scoring.

How it works

Understand the prompt evolution lifecycle, evaluation pipeline, and scoring mechanics.

API Reference

Explore the Bitrecs V2 REST API for submissions, scoring, and inference.

​
What is Bitrecs?

Bitrecs is an ecommerce recommendation engine powered by Bittensor. Merchants use the Bitrecs widget to drive personalized customer journeys and increase average order values. The quality of recommendations is continuously improved by miners competing to optimize LLM prompts through the Bitrecs V2 subnet.
1

Acquire a hotkey

Register a fresh hotkey on the Bitrecs subnet (Netuid 122 for mainnet, 296 for testnet) using btcli.
2

Create an artifact

Author an artifact.yaml containing your system prompt, user prompt, model choice, and sampling parameters. Upload it as a single-commit GitHub Gist.
3

Submit via CLI

Run the Bitrecs CLI to make an onchain commitment and submit your artifact for evaluation.
uv run bitrecs_cli.py upload \
  --github-account your-account \
  --gist-id your-gist-id \
  --coldkey-name default \
  --hotkey-name default
4

Earn emissions

Your artifact enters the evaluation pipeline — two screeners and multiple validator runs. If your artifact dominates the Pareto frontier and wins the WTA scoring round, you earn onchain emissions.

​
Key concepts

Artifact reference

Learn every field in the artifact.yaml format, including required fields and template variables.

Scoring mechanics

Understand how Pareto dominance, WTA scoring, and time-decay combine to determine emissions.

Submission process

Step-by-step guide to onchain commitment, validation, and artifact submission.

Architecture

See how the API, validators, evaluators, and Bittensor network fit together.