## What Mira Network Does
- Mira transforms AI or human-generated content into smaller factual “claims” and sends them to a distributed network of verifier models and nodes to check correctness.
- The network aggregates these checks, reaches consensus, and issues a cryptographic certificate that proves which claims are true and which models agreed.
- Its goal is to reduce AI “hallucinations” and bias and move real-world AI usage (medicine, law, finance, etc.) from roughly 70% to 95%+ reliability by adding a trust layer on top of any model.
## Tech & Architecture Fundamentals
- **Core idea:** trustless AI output verification via “content → claims → distributed verification → on-chain or signed certificate.”
- Claims are processed by independent verifier nodes running different AI models; multiple consensus modes (e.g., N-of-M) can be used depending on the use case.
- Mira runs as an ERC‑20 token on Base (Ethereum L2), giving it EVM compatibility and DeFi composability.
- The broader stack includes a Mira Blockchain / settlement layer, Model Hub, Knowledge Market, and Flows marketplace for AI “flows” (composable AI pipelines), all coordinated via the Mira SDK.
## Token Utility & Economics
- MIRA is the **utility + governance** token for the whole ecosystem: it pays for API access, verification calls, and SDK usage, and is used for staking and protocol decisions.
- Node operators must stake MIRA to participate in verification; malicious or low-quality behavior can be punished via slashing, while honest verification earns network fees and rewards.
- All platform usage (Verified Generate API, Flows marketplace, other AI services) requires MIRA payments, with benefits (priority access, better pricing) for holders, creating direct demand tied to actual usage.
- MIRA also acts as the base pair asset for ecosystem tokens, so new apps that launch their own tokens often need MIRA for liquidity pairing and routing.
### Tokenomics (Supply & Allocation)
- Max supply: 1 billion MIRA.[8]
- Allocation example: 26% ecosystem fund, 20% core team, 16% node rewards, 15% foundation treasury, 14% early investors, 6% initial airdrop, 3% liquidity incentives.
- Emissions and key parameters are governed by token holders, who vote on things like emission schedules, network upgrades, and strategic design changes.
## Example Use Cases
- Verifying outputs from LLMs before they hit production in sectors like finance, healthcare, or legal, with a proof attached to each response.
- Powering apps like Delphi Oracle, Astro, and Amor, which use Mira’s verification layer as infrastructure for reliable AI insights or actions.
- Enabling marketplaces (Model Hub, Knowledge Market, Flows Market) where compute providers, model authors, and data owners get streamed MIRA payouts for usage of their resources.
## Key Strengths and Risks
**Strengths**
- Clear real-world problem: AI reliability and hallucination.
- Strong crypto-economic design (staking + slashing, node rewards, fee-based demand).
- Infrastructure role: potential to become a “trust layer” used by many AI apps rather than a single app bet.
**Risks**
- Competes with other AI+crypto stacks and centralized AI guardrail solutions.
- Adoption risk: value heavily depends on how many real apps route their verification through Mira.
- Usual crypto risks: volatility, regulatory environment, execution risk of the team and roadmap.
If you want, I can next break this down specifically from an investor’s angle (valuation drivers, what metrics to track, and where it might fit in a crypto AI portfolio).