primer · flock.io

Four FLock kernels, demystified.

Every idea in this repo leans on one of four FLock.io kernels. Each one runs against the public FLock API Platform (https://api.flock.io/v1, OpenAI-compatible) with a single key, so you ship a real demo with zero infra and no provider-shopping.

Why FLock.io

FLock.io is a decentralised-AI project (Base chain, FLOCK token, AI Arena for training, FL Alliance for federated fine-tuning). For this hackathon we use just one slice — the API Platform, an enterprise-grade model hub that fronts the whole marketplace with one OpenAI-compatible endpoint and significantly lower inference cost than mainstream APIs.

frontier reasoning

FLock Chat (Text Models)

Kernel: an OpenAI-compatible `POST /v1/chat/completions` against https://api.flock.io/v1 — pick any FLock marketplace model (DeepSeek V4 Flash/Pro, Gemini 3.5 Flash, Kimi K2.6, Qwen3, MiniMax) through one endpoint, one key

UI: a prompt box or chat surface that returns markdown the user can read, copy, or act on

visual generation

FLock Image Generation

Kernel: the FLock marketplace image models (Nano Banana 2, Gemini 3.1 Pro Preview, Gemini 3 Flash) called via chat completions with `modalities: ['image','text']` on https://api.flock.io/v1 — returns a hosted image URL the client renders straight into an <img>

UI: a prompt-driven canvas that renders covers, moodboards, posters or scene art on demand

motion generation

FLock Video Generation

Kernel: the FLock video-capable models (Kimi K2.5) called on https://api.flock.io/v1 — async task with polling, returns an MP4 URL the client renders as a short reel

UI: a prompt → progress bar → playable MP4 surface for short motion sketches, loops, or scene reels

function calling

FLock Agents (Tool Use)

Kernel: FLock's tool-capable models (DeepSeek V4 Pro, Kimi K2.6, Gemini 3 Flash) with OpenAI-style `tools` on `POST /v1/chat/completions` — the model plans, calls typed tools, and returns a grounded answer

UI: a task box that returns a structured answer or plan the user can follow, with visible tool calls

One secret. That's it.

Add this to your Lovable project (Project Settings → Secrets). It stays on the server, read by your TanStack server function via process.env.FLOCK_API_KEY and sent to FLock in the x-litellm-api-key header.

Build strategyBrowse 1,000 ideas