Frequently asked questions

FLock.io is a decentralised AI training platform using federated learning and blockchain verification. Enterprises and communities can build privacy-preserving models without exposing raw data. Core modules include AI Arena, FL Alliance, the API Platform, and Moonbase-FOMO. Docs: https://docs.flock.io/

FLock is designed so organisations and communities keep data local during collaborative training. Models improve without moving or exposing underlying private datasets. Federated learning plus on-chain verification supports trust-minimised coordination across participants.

AI Arena is FLock.io’s decentralised AI training platform, accessed through train.flock.io. Participants stake FLOCK on Base to join and earn staking rewards. Roles include task creators, training nodes, validators, and delegators.

AI Arena supports task creators who define desired models and submit tasks. As documented, tasks are currently created by the FLock.io team. FLock states community task creation will open in coming months.

AI Arena has four participation roles: task creators, training nodes, validators, and delegators. Training nodes develop or fine-tune models. Validators evaluate submissions. Delegators stake to training nodes and validators to increase reward distribution.

FL Alliance further refines model parameters after AI Arena optimisation, using proprietary local data in a federated learning setting. FLock documents an incentive design to reward honest behaviour and penalise malicious FL participants.

AI Arena focuses on decentralised training and validation of submitted models. FL Alliance follows with federated fine-tuning on proprietary data. Together they form a pipeline from open competitive training toward private collaborative refinement.

The FLock API Platform is the deployment layer for using FLock-trained models in real applications. FLock positions it as infrastructure for bringing co-created models into products through an API-oriented interface on platform.flock.io.

FLock.io uses FLOCK, its token deployed on Base. In AI Arena, users stake FLOCK through train.flock.io to participate and earn staking rewards tied to training, validation, or delegation activity.

FLock combines federated learning for local data training with blockchain-based verification and incentives. Docs describe blockchain as the incentivisation and coordination layer so contributions, rewards, and participation rules can be enforced without a single trusted operator.

Yes for node participation workflows described in docs: a private key initialises the client, authenticates the user, and validates on-chain transactions tied to contributions. Training runs locally; FLock docs state they do not store user private keys on a central server.

FLock documentation states training operations run locally and private keys are not exposed to FLock or stored on a central server. Users remain responsible for safeguarding their own keys.

On flock.io, FLock names collaborations including the United Nations Development Programme (UNDP), Alibaba Cloud’s Qwen, and the Ethereum Foundation, among other ecosystem partners quoted on the homepage. Treat homepage naming as marketing attribution, not a legal partnership register.

Official docs live at https://docs.flock.io/, including product pages for AI Arena and FL Alliance, system design, and resources. Markdown versions are available by appending .md to docs URLs.

The public contact email listed on flock.io is hello@flock.io. For product help, start with docs.flock.io and community channels linked from the site footer.