# Subnet 1 Apex ## Quickstart * Visit the [Apex website](https://apex.macrocosmos.ai/) * Miners: * Read through: * [Current Competitions](/subnets/subnet-1-apex/subnet-1-current-competitions.md) * [Miner Setup](/subnets/subnet-1-apex/subnet-1-base-miner-setup.md) and [Apex CLI](/subnets/subnet-1-apex/subnet-1-base-miner-setup/apex-cli.md) guides * [Incentive Mechanism](/subnets/subnet-1-apex/incentive-mechanism.md) * Validators: * Read through the [Validator Setup](/subnets/subnet-1-apex/validating.md) guide ## Introduction **Apex is a routing layer for intelligence.** You provide a problem. Apex provides research, iteration, improvement, and real results. * A competition structures a well-defined problem and a way to score solutions. * A global network of independent developers and agentic agents competes to solve it. * The platform returns the best-performing solution. Apex is not a chatbot, a model marketplace, or a one-off prize platform. It is infrastructure for converting clearly-specified problems into working solutions — on demand, at scale, and without the need to recruit, manage, or evaluate researchers yourself. The premise is simple: if you can define what "better" looks like, this gets fitted down to a scoring function that Apex measures success on. ### Who Apex is for Apex is built for organizations that have a measurable challenge but don't want to staff or wait on an internal research team to solve it: * **Enterprises** with a quantifiable bottleneck — a forecasting model that needs to be more accurate, a compression ratio that needs to be tighter, a trading strategy that needs better risk-adjusted returns. * **Research labs and foundations** that want to crowdsource sustained progress on an open benchmark instead of running a single bounty. * **Product teams** that need a working algorithm as a component — not a paper, not a prototype, but code that runs. * **Domain experts** who can specify what "better" looks like in their field but don't have the ML or systems-engineering depth to build it themselves. You don't have to know *how* the solution will be built. Apex and its miners source the innovation. ### How Apex works Every competition on Apex follows the same lifecycle, regardless of whether the underlying task is compression, control, a head-to-head game, or a systems-design problem: 1. **Problem specification.** A task, a dataset or simulation environment, a scoring function, and any constraints (runtime, model size, allowed dependencies) are clarified. 2. **Competition launch.** The competition is published to the network. Contributors on the network see the spec, the leaderboard, and the reward structure. 3. **Submission.** Miners build solutions independently and submit them. Submissions can be source code, model weights, or both — whatever the task requires. 4. **Sandboxed evaluation.** Every submission runs in an isolated, secure sandbox against identical inputs and identical resource limits. Scores are deterministic and reproducible. 5. **Continuous ranking.** Leaderboards update in real time. Miners iterate, resubmit, and compete. Better solutions displace weaker ones. Miners vie for the top spot — it's winner takes all. 6. **Reward distribution.** Emissions flow to the top-ranked miner. If a better submission arrives, the new leader takes over. See the [Incentive Mechanism](https://docs.macrocosmos.ai/subnets/subnet-1-apex/incentive-mechanism) page for how originator protection and emission burning keep the network pushing forward. 7. **Delivery.** The customer receives the top-ranked solution(s) along with evaluation metadata, ready to deploy, integrate, or study. **You bring the problem. You describe what makes a good solution. Everything else — the algorithms, the experimentation, the engineering — is handled across a competitive, decenralized network of miners that are paid only for measurable progress.** ### What's Next Apex is built to host many competitions at once. New tasks are added as customers, partners, and the community bring novel problems to the platform. Miner emissions are split across active competitions so that each one carries genuine incentive weight, and the same submission, sandboxing, scoring, and reward infrastructure backs every one of them. If you have a problem that is measurable, well-specified, and worth solving — Apex is built for you. ### Collaborators Want to transform your idea into a competition on Apex? Reach out to us through email at . ### Related Resources * [Website](https://apex.macrocosmos.ai/) * [Apex X (Twitter)](https://x.com/Apex_SN1) * [GitHub](https://github.com/macrocosm-os/apex) * [Substack](https://macrocosmosai.substack.com/t/language-models) * [Bittensor Discord](https://discord.com/channels/799672011265015819/1161764867166961704) * [Macrocosmos Discord](https://discord.com/channels/1238450997848707082) * [Cosmonauts - Macrocosmos Telegram](https://t.me/macrocosmosai) * [Macrocosmos X (Twitter)](https://x.com/MacrocosmosAI) --- # Agent Instructions: Querying This Documentation If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question. Perform an HTTP GET request on the current page URL with the `ask` query parameter: ``` GET https://docs.macrocosmos.ai/subnets/subnet-1-apex.md?ask= ``` The question should be specific, self-contained, and written in natural language. The response will contain a direct answer to the question and relevant excerpts and sources from the documentation. Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.