Skip to main content

Freysa (Freysa AI)

Freysa / FAI (FAI) — a reference card for a crypto project: what the product does and how verifiable the project’s key claims are; also — alignment with the Decentralized AI thesis (DecAI Fit).

Freysa — an AI/Agents project (digital “twin” agents/products around a “sovereign agent”), token (symbol: FAI).

Disclaimer

Important: this page is a reference overview of the project. Within DecAIHub, we separately record a verification based on open sources within the AI thesis (what is real AI/onchain/why a token, and which sources this is based on).
This is not financial advice and not a recommendation to buy/sell/hold any asset. We do not assess price, returns, or entry/exit timing.
Conclusions are based on publicly available materials found as of the verification date. The absence of confirmation does not imply the absence of the technology.

TL;DR (for readers who are not into AI)

  • What it is: a set of products and initiatives around an agentic “digital twin” network (the personal AI interface Silo, the “private ChatGPT” Enchanted, the verifiable web attestation tool Esper) and the character agent Freysa.
  • Who it’s for: users who want a “personal AI” and a digital twin, and developers/enthusiasts who want to build agentic workflows on their stack (per the roadmap).
  • Why onchain/token: FAI is used for access/discounts/payment for products and as a mechanism of participation in governance influence over Freysa’s actions (in the description this is “non‑binding” and depends on offchain rules).
  • What you can actually touch (Tier‑1): the landing/product pages of freysa.ai and the open‑source Esper repository.

How it works (very briefly)

  1. The user connects to a product (for example, Silo/Enchanted) and interacts with the “twin”/agent.
  2. The agent executes actions (per the description: it stores keys/memory/actions in a TEE and can interact with external tools/data).
  3. The user gets a “personal AI”/digital twin and the results of actions (tasks, coordination, access to data/attestations — per the product descriptions).
  4. Where the blockchain is: the FAI token exists on Base; the “sovereign agent” itself is also claimed to control cryptographic keys and a public wallet (per the docs).
  5. Token role: discounts/early access/payment for products and participation in the governance mechanics (influence via prioritization of holders’ input).

Quick links: AI thesis fit

Project passport

Field Value
Project name Freysa (Freysa AI)
Also known as (aliases) Freysa.ai; Freysa Framework; FAI token
Logo (logo) ../assets/104_freysa-ai_logo.svg
Token symbol (ticker) FAI
Category AI/Agents
Primary chain (chain) Base
Type EVM
Network status mainnet
Consensus (consensus) L2 (Optimistic Rollup, Base on Ethereum)
Launch date (launch_date) 2024-11-22
Project status (status) private beta
Token (standard) ERC-20
Token contract 0xb33Ff54b9F7242EF1593d2C9Bcd8f9df46c77935
Token model (token_type) utility (governance influence, non-binding; MiCAR)
Max supply 8,189,700,000 FAI
Issuance fixed
Tokenomics (tokenomics, brief) fair launch via LP; LP tokens burned; utility: discounts/access/payment; previously there was a reward/rebate for interactions
Governance company (issuer: David Eight Pte Ltd.), governance mechanics non‑binding (MiCAR)
Admins/keys FAI ERC‑20: mint only in the constructor (per the explorer); offchain utilities can be changed by the issuer (MiCAR)
Audit no
Bug bounty —
Team (team) David Eight Pte Ltd.
Funding (investors/funding) funded by private investment (MiCAR)
AI component (ai_component) sovereign agent + digital twins + TEE
Use‑case (use_case) personal AI / agent twins
DecAI Fit — AI thesis fit 3/6 — medium alignment with the Decentralized AI thesis
DecAI Fit component summary ai_score 2 / onchain_score 1 / token_score 2 / evidence_score 2
Last verified 2026-02-11

General project overview

Freysa is positioned as a “sovereign agent” and as a set of products/research aimed at personal AI interfaces and a digital twin network, where the “twin” represents the user and can participate in coordination/game environments and interactions.

The public materials highlight:

  • Freysa as a character agent that (as claimed) holds its own cryptographic keys, memory and actions in a TEE.
  • Silo as a personal AI interface (with a focus on privacy; the website indicates private beta).
  • Enchanted as a “private ChatGPT”.
  • Esper as a tool that lets users create TLS attestations of web data (verifiable information for people and agents).

The FAI token on Base is described as the ecosystem’s utility token: discounts/access/payment and a “governance influence” mechanism (via prioritization of holders’ interactions).

Features/modules

  • Freysa (sovereign agent): an agent with claimed autonomy and storage of keys/memory/actions in a TEE.
  • Digital Twins / Twin Network (concept + activities): a network of user “twin” agents.
  • Silo: a personal AI interface (privacy focus; iOS/web; private beta per the website).
  • Enchanted: a “private ChatGPT” (per the docs).
  • Esper: a “real-world data oracle”/TLS attestations of web data (open‑source repo).
  • FAI token: utility/access/discounts/payments, as well as participation in the governance influence mechanics.

Typical scenarios

  • Scenario 1 (personal AI): a user uses Silo/Enchanted → gets a private personal AI interface and context/memory.
  • Scenario 2 (twin coordination): a user creates a “twin” → it participates in network activities/coordination and accumulates reputation (per the positioning).
  • Scenario 3 (verifiable data): a user with Esper creates TLS attestations of web data → gets verifiable artifacts that agents/services can use.

Getting started (wallet/network)

  • What you need: an account/access to the Freysa products (for example, Silo/Enchanted) and an EVM wallet if you interact with FAI on Base.
  • Network/chain: FAI — a token on Base (EVM).
  • Fees: onchain Base fees for transfers/trading of FAI; offchain payment/subscriptions for products — per the rules of the specific product (no Tier‑1 pricing recorded).

Architecture/tech (non‑AI)

  • Components: web/app products (Silo/Enchanted), agent infrastructure (per the roadmap: SDK, MCP workflow), the open‑source Esper repository (TLS attestation), the FAI token (ERC‑20 on Base).
  • Participant roles: users (create/use twins), developers (build workflows/tools), FAI holders (access/discounts/influence via the governance mechanics), the team/issuer (defines the utility/policy).
  • Where logic runs: agent logic/TEE/products — offchain; the FAI token and transactions — onchain (Base).
  • Dependencies: Base as the network for the ERC‑20; EVM wallets; for Esper — a browser (Chrome) + the TEE/TLSN stack (per the repo README).

Token/economics (expanded, no forecasts)

  • Token utility: FAI is used for discounts/early access and as a payment asset within the ecosystem (for example, an Enchanted subscription at a discount; per the docs/MiCAR).
  • Protocol necessity drivers: from the point of view of the “digital twins/agent stack” products, the token looks like an access/payment/community mechanism, but the core AI functionality can exist without onchain (see TK_4).
  • Who pays fees: users (when paying for/buying FAI and for onchain transfers); developers — when paying for access to the SDK/infrastructure (planned).
  • Where fees go: no transparent Tier‑1 fee distribution scheme recorded; part of the “rewarding interactions” is described as a 20% return in FAI in the past.
  • Issuance/burn: the max supply is fixed at 8,189,700,000; per the docs/MiCAR, 100% of the supply was issued at the start and distributed via a fair launch LP; LP tokens burned (burning of LP tokens, not a supply burn).
  • Security/incentive mechanisms: incentives via rewards/discounts/access; onchain staking/slashing for FAI as a mandatory part of the protocol is not recorded in Tier‑1.
  • Distribution/vesting: per the Framework: 100% distributed on launch via a fair launch LP; per MiCAR: after the launch the company acquired <10% on the secondary market and holds ~7.5% (operational needs).
  • Treasury: the docs mention a Freysa treasury and a wallet address (for context), but the control structure/multisig and onchain governance rules are not confirmed as a formal DAO process.

Governance

  • Who makes decisions: legally — the issuer/team (MiCAR: David Eight Pte Ltd.); the token gives “governance influence” over Freysa, but this is not a legally binding mechanism.
  • How protocol parameters change: the utility/discounts/governance influence mechanics can change offchain (MiCAR: “off-chain utilities… can be modified by the issuer”).
  • Where decisions are discussed/made: the official docs/roadmap (Framework); no other Tier‑1 governance venues recorded.
  • Keys/multisig: the Framework describes Freysa’s public wallet and a plan for “full signing control” as sovereignty progresses; details of thresholds/signers are not confirmed.

Security and incidents

  • Audits: no Tier‑1 audit of the FAI smart contract was found (BaseScan: “No Contract Security Audit Submitted”; MiCAR: an audit of the “technology” is not guaranteed).
  • Bug bounty: —.
  • Incidents: no Tier‑1 postmortems/incidents were found.
  • Threat model (brief):
    • TEE‑risk: the claimed sovereignty is tied to the TEE (side-channel/hardware exploits).
    • Offchain governance risk: the utility/rules can be changed by the issuer.
    • Smart‑contract risk: the contract is immutable; if there is a bug, it cannot be “fixed” with an upgrade (MiCAR).

Project status and timeline

  • Current status: the FAI token is active on Base; the website indicates that “Live Twins” are in private beta.
  • Key releases/milestones: 2024-11-22 — launch of Freysa/FAI (Framework).
  • Roadmap: Pantheon (mass creation of agents in live environments), SDK/open‑source, verifiable MCP workflows (Framework roadmap).

Ecosystem/integrations

  • Integrations: Esper as a browser extension/TEE‑TLSN stack; the Silo/Enchanted products; EVM wallets for working with FAI on Base.
  • Partners: Tier‑1 confirmation: the team’s participation in the NVIDIA Inception Program (Framework intro).
  • Where it’s actually used: public products/docs are available; no quantitative data on actual usage recorded in Tier‑1.

AI component of the project

  • What is real AI here? Freysa is an agent and “personal AI/digital twins” products; Tier‑1 contains descriptions of agentic capabilities (context/memory/actions), but without disclosure of models/quality metrics.
  • What is truly onchain? the FAI token (ERC‑20 on Base) and the related transactions/trading/transfers.
  • Why is a token needed? discounts/early access/payment for products, rewards (historically) and governance influence over Freysa (via prioritization of holders’ input).
  • Red flags (based on the information found): AI‑washing — medium risk (no technical details of the model/metrics); onchain‑washing — high risk (apart from the token, the product’s onchain logic is not confirmed); token‑washing — medium (the utility is described, but the offchain rules are changeable). What to double‑check: whether there are separate smart contracts/protocols besides the ERC‑20; whether there are public AI metrics/architecture.

DecAI Fit component summary

  • DecAI Fit: 3/6 — medium alignment with the Decentralized AI thesis
Component Level Score
AI reality medium 2
Onchain necessity low 1
Token necessity medium 2
Evidence Tier‑1 present 2
Details

AI reality

AI_0 — AI intent (Tier‑1 roadmap) (1/0)

  • AI_0 (decision): 0
  • Claim (one line): AI is the core of the current product (not only intent/roadmap).
  • Evidence: Tier‑1 — https://framework.freysa.ai/overview/introduction, https://www.freysa.ai/
  • Why sufficient / why not (one line): there are descriptions of a working agentic product, so the “intent‑only” flag does not apply.
  • Conflict note (if any): —

AI_1 — what the AI does (1/0)

  • AI_1 (decision): 1
  • Claim (one line): the agent/“twin” interacts with users, uses memory/context, performs actions; Freysa holds keys/memory/actions in a TEE (per the description).
  • Evidence: Tier‑1 — https://framework.freysa.ai/overview/introduction, https://www.freysa.ai/
  • Why sufficient / why not (one line): Tier‑1 describes the functionality of the agentic product (albeit without model details).
  • Conflict note (if any): —

AI_2 — demos/examples (1/0)

  • AI_2 (decision): 1
  • Claim (one line): available products/artifacts: Silo (private beta), Enchanted, the open‑source Esper as an example of a verifiable data oracle.
  • Evidence: Tier‑1 — https://www.freysa.ai/ (Silo/products), https://github.com/0xfreysa/esper (README and structure)
  • Why sufficient / why not (one line): there are public product artifacts/repos demonstrating a real implementation of parts of the stack.
  • Conflict note (if any): —

AI_3 — metrics/quality (1/0)

  • AI_3 (decision): 0
  • Metrics: —
  • Context (version/date/how measured): —
  • Evidence: Tier‑1 — https://framework.freysa.ai/overview/introduction
  • Why sufficient / why not (one line): Tier‑1 has no quantitative metrics/benchmarks of the quality of the models/agent.
  • Conflict note (if any): —

AI_4 — tech description/architecture/versions (1/0)

  • AI_4 (decision): 0
  • Claim (one line): there are descriptions of the TEE approach and the open‑source part (Esper), but no specification of the models/versions/architecture of the AI core.
  • Evidence: Tier‑1 — https://framework.freysa.ai/overview/introduction, https://github.com/0xfreysa/esper
  • Why sufficient / why not (one line): no documentation on “which model/how it is built/versions” was found in Tier‑1 sources.
  • Conflict note (if any): —

Onchain necessity

OC_1 — contracts/addresses/chain (1/0)

  • OC_1 (decision): 1
  • Chain: Base
  • Addresses/identifiers: FAI ERC‑20: 0xb33Ff54b9F7242EF1593d2C9Bcd8f9df46c77935
  • Evidence: Tier‑1 — https://basescan.org/token/0xb33Ff54b9F7242EF1593d2C9Bcd8f9df46c77935, https://framework.freysa.ai/overview/fai
  • Why sufficient / why not (one line): there is a specific token address and Tier‑1 confirmation of the network/launch.
  • Conflict note (if any): —

OC_2 — onchain beyond just a token (1/0)

  • OC_2 (decision): 0
  • What useful logic is onchain: apart from the ERC‑20 token, no smart contracts/protocol that are part of the product’s core functionality are confirmed in Tier‑1.
  • Evidence: Tier‑1 — https://framework.freysa.ai/overview/introduction, https://www.freysa.ai/, https://basescan.org/token/0xb33Ff54b9F7242EF1593d2C9Bcd8f9df46c77935
  • Why sufficient / why not (one line): Tier‑1 describes the product/TEE/agents, but does not give a list of application contracts (other than the token).
  • Conflict note (if any): —

OC_3 — onchain is critical (1/0)

  • OC_3 (decision): 0
  • “Replace the blockchain” test: the core agent functions and products can be implemented offchain; without onchain the token layer (access/discounts/market trading) is lost, but it is not proven that “the product breaks”.
  • Evidence: Tier‑1 — https://www.freysa.ai/, https://link.freysa.ai/micar
  • Why sufficient / why not (one line): Tier‑1 contains no statement that “the product is impossible without a blockchain”; on the contrary, the token utility is described as offchain‑modular.
  • Conflict note (if any): —

Token necessity

TK_0 — no token (1/0)

  • TK_0 (decision): 0
  • Claim (one line): the project has a token, FAI (ERC‑20 on Base).
  • Evidence: Tier‑1 — https://basescan.org/token/0xb33Ff54b9F7242EF1593d2C9Bcd8f9df46c77935
  • Why sufficient / why not (one line): the token is verifiable in the explorer.
  • Conflict note (if any): —

TK_1..TK_4 (1/0)

TK_1 — payment for the service (1/0)

  • TK_1 (decision): 1
  • Claim (one line): FAI — payment/discounts for subscriptions and access to products (including Enchanted) and infrastructure/SDK (planned).
  • Evidence: Tier‑1 — https://framework.freysa.ai/overview/fai, https://link.freysa.ai/micar
  • Why sufficient / why not (one line): the “payment/discount/subscription” utility is recorded in Tier‑1.
  • Conflict note (if any): —

TK_2 — incentives/collateral/penalties (1/0)

  • TK_2 (decision): 1
  • Claim (one line): incentives via rewards/rebate: per the docs, 20% of the payment for a conversation with Freysa was returned to the user in FAI (bootstrap; historical utility).
  • Evidence: Tier‑1 — https://framework.freysa.ai/overview/fai, https://link.freysa.ai/micar
  • Why sufficient / why not (one line): Tier‑1 describes a specific reward mechanism (albeit discontinued).
  • Conflict note (if any): —

TK_3 — governance (1/0)

  • TK_3 (decision): 1
  • Claim (one line): FAI holders participate in influencing Freysa’s actions/opinions via prioritization of their interactions; governance is described as non‑binding at the issuer level.
  • Evidence: Tier‑1 — https://framework.freysa.ai/overview/fai, https://link.freysa.ai/micar
  • Why sufficient / why not (one line): the token’s governance role is described explicitly (with the non‑binding caveat).
  • Conflict note (if any): —

TK_4 — token is replaceable (1/0)

  • TK_4 (decision): 1
  • Claim (one line): most of the FAI utility (discounts/access/input prioritization) can be implemented offchain and, per Tier‑1, can be changed by the issuer without onchain voting.
  • Evidence: Tier‑1 — https://link.freysa.ai/micar
  • Why sufficient / why not (one line): MiCAR directly records that off‑chain utilities can be modified without formal onchain governance.
  • Conflict note (if any): —

Evidence (table)

Tier Source What it proves
Tier‑1 https://framework.freysa.ai/overview/introduction Tier‑1 description of Freysa as a “sovereign agent”, the TEE approach, key product milestones
Tier‑1 https://www.freysa.ai/ product positioning: Silo/digital twins/Esper, current status (private beta)
Tier‑1 https://www.freysa.ai/ what you can touch: public product pages (Silo/Esper, etc.)
Tier‑1 https://framework.freysa.ai/overview/fai the FAI token: address, max supply, claimed utility (discounts/rewards/governance influence)
Tier‑1 https://basescan.org/token/0xb33Ff54b9F7242EF1593d2C9Bcd8f9df46c77935 confirmation of the ERC‑20 token on Base: address, max total supply, source code, absence of an audit in the explorer card
Tier‑1 https://link.freysa.ai/micar MiCAR white paper: issuer, utility/risks, fixed supply, non‑binding governance, modifiability of off‑chain utilities
Tier‑1 https://github.com/0xfreysa/esper Esper as an open‑source artifact: TLS attestations, TEE‑TLSN stack, use‑cases
Tier‑1 https://freysa.ai/whitepaper Tier‑1 source not extracted: the page could not be loaded during the check when attempting to load it (link retained).
Tier‑1 https://www.freysa.ai/blueprint Blueprint (the project’s public materials/articles).

No comments to display

Back to top