# inference-labs-inc/subnet-2

Verifiable inference on Bittensor

Repository: https://github.com/inference-labs-inc/subnet-2
Canonical: https://ross.abutalabs.com/products/subnet-2
Homepage: https://subnet2.inferencelabs.com
Language: Rust
License: MIT
License Family: permissive
Topics: bittensor, zk, zkml
Last push: 2026-08-12T15:33:40+00:00

## Health v2 (maintenance only)
Score: 86/100 (v2, computed 2026-09-03T02:39:23.370411+00:00)
- activity 97, release rhythm 85, longevity 62
- inputs: {"age_days": 875, "days_push": 21, "days_rel": 23, "gap_med": 0, "n_releases_24m": 262}
- flags: none
- formula: round(0.45*activity + 0.35*rhythm + 0.20*longevity); archived -> min(score, 10)

## Adoption (not part of the score)
Stars 2437, forks 127 (observed 2026-08-28T04:06:51.464378+00:00)

## What it is
Subnet 2 on the Bittensor network implements a Proof-of-Inference system that uses zero-knowledge machine learning (zk-ML) to cryptographically verify that AI predictions were produced by a specific model. It ships native Rust miner and validator binaries that communicate over HTTP and QUIC, with pre-built auto-updating releases.

## Use cases
- verify that AI model predictions come from a specific model using zero-knowledge proofs
- run a miner that serves zk-circuit-verifiable AI predictions on Bittensor
- run a validator that checks zk-proofs of inference from miners
- earn rewards for generating succinct, efficient zk-provable models
- participate in a decentralized verified intelligence network without GPU hardware
- audit the integrity of AI inference outputs cryptographically

## When to choose
- you want to mine or validate on Bittensor subnet 2 with verifiable inference
- you need cryptographic proof that an AI prediction came from a specific model
- you want to contribute to or build on zk-ML proving systems in a decentralized network
- you prefer CPU-friendly participation since zk-proofs are CPU-intensive

## When to avoid
- you need a general-purpose zk-ML toolkit outside the Bittensor ecosystem
- you want a simple hosted inference API without blockchain participation
- you are not prepared to operate continuously running miner/validator nodes with auto-updates

## Facets
- artifact type: service
- maturity: active
- function: machine-learning, llm-inference, cryptography, monitoring, cli
- domain: machine-learning, blockchain, artificial-intelligence, security, microservices
- platform: rust, cli
- tags: zkml, bittensor, proof-of-inference, zero-knowledge-proofs, subnet, verifiable-inference, decentralized-ai, miner, validator, cryptocurrency, linux, macos

## Member repositories
- inference-labs-inc/subnet-2 (main) score 86

## Provenance
- Observed fields: from GitHub, fetched 2026-08-28T04:06:51.464378+00:00.
- Health v2: computed from the inputs above; adoption is never an input.
- Inferred fields (summary, facets, guidance): AI-extracted, prompt v1, taxonomy v1, on 2026-08-30T02:31:00.441873+00:00, confidence not recorded.
  - readme: https://github.com/inference-labs-inc/subnet-2 (fetched 2026-08-28T04:06:51.464378+00:00, sha 50338ffacbc0)
  - homepage: https://subnet2.inferencelabs.com (fetched 2026-08-29T10:12:18.081104+00:00, sha 46f6793332dc)
- Data as of 2026-08-30T08:39:29.467469+00:00.