Skip to main content

Cortex

Cortex (CTXC) — 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).

Cortex — an AI/Inference / L1 / Infra blockchain platform that claims deterministic onchain inference of ML models in its EVM‑compatible virtual machine (CVM) using GPUs and a deterministic inference engine (Synapse), as well as an economy of rewarding model providers through fees (endorphin).

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: an L1 blockchain where smart contracts can call ML models and receive the result onchain (with consensus verification of the computations).
  • Who it’s for: developers of AI smart contracts and dApps; AI developers who publish models; miners/nodes.
  • Why onchain/token: the blockchain is needed for consensus/verifiability and the distribution of rewards; CTXC is used to pay execution fees (endorphin) and rewards to miners/model providers.
  • What you can actually touch (Tier‑1): https://cerebro.cortexlabs.ai/

How it works (very briefly)

  1. The user does: deploys a smart contract (Solidity) and calls a function that invokes an onchain model via infer/inferArray.
  2. The system does: the CVM (EVM‑compatible) executes the contract and runs the deterministic inference engine (Synapse) for the model stored in the Cortex storage layer; the computation is deterministic so that it can be validated by consensus.
  3. The user gets: the inference result within the transaction/contract (the onchain logic can react to it).
  4. Where the blockchain is: storage/addresses of models and the execution of inference + the rules for fee payment/distribution happen onchain.
  5. Token role (if any): CTXC pays the endorphin fees for operations; rewards are distributed to miners and model providers (per the docs).

Quick links: AI thesis fit

Project passport

Field Value
Project name Cortex
Also known as (aliases) Cortex Labs; Cortex Blockchain; Cortex Virtual Machine (CVM); Synapse (deterministic inference engine)
Logo (logo) ../assets/214_cortex_logo.png
Token symbol (ticker) CTXC
Category AI/Inference | L1 | Infra
Primary chain (chain) Cortex
Type non-EVM
Network status mainnet
Consensus (consensus) PoW (Cuckoo Cycle / CuckARoo 30), target block time ~15s
Launch date (launch_date) unknown (the extracted Tier‑1 sources contain no exact mainnet launch date)
Project status (status) active (docs, the CortexTheseus client, roadmap/updates on the website)
Token (standard) native
Token contract — (native coin of the Cortex network)
Token model (token_type) utility (fee token + mining rewards + model-provider rewards)
Max supply total supply 299,792,458 CTXC (per the mining docs)
Issuance PoW mining: block reward 7 CTXC, halving every ~4 years (per the mining docs)
Tokenomics (tokenomics, brief) fees (“endorphins”) are paid in CTXC and distributed between the miner and the AI model provider (per the docs)
Governance unknown
Admins/keys unknown (L1 PoW; protocol parameters are managed via client updates)
Audit unknown (the extracted Tier‑1 contains no public audit report)
Bug bounty yes (bug bounty program in the tech-doc)
Team (team) Cortex Foundation / Cortex Labs (per the repositories and the website; the roster not extracted)
Funding (investors/funding) unknown
AI component (ai_component) deterministic on-chain ML inference (Synapse/CVM)
Use‑case (use_case) AI smart contracts + AI dApps
DecAI Fit — AI thesis fit 5/6 — high alignment with the Decentralized AI thesis
DecAI Fit component summary 3 / 3 / 1 / 3
Last verified 2026-02-11

General project overview

Cortex is an L1 blockchain network with a modified EVM (CVM) that claims the ability to perform machine learning onchain: a smart contract can call a model stored in the network, run inference and use the result in onchain logic. Unlike ordinary dApps, which are forced to do inference offchain, Cortex emphasizes the determinism of computations, so that the result can be agreed upon/verified by consensus.

The Cortex tech documentation describes: the CVM (EVM‑compatible + infer instructions), the deterministic inference engine Synapse, the MRT mechanism (deterministic quantization/integer inference) and the Endorphin fee model, in which part of the reward is directed to model providers, incentivizing the publication of models to the network.

Features/modules

  • Cortex Virtual Machine (CVM): an EVM‑compatible VM with support for AI inference and the infer/inferArray instructions.
  • Synapse: a deterministic inference engine that ensures the same result across different computing environments.
  • MRT: a deterministic fixed‑point/quantization approach so that models are executable in the CVM.
  • Storage layer for models/data: publication of models/data to the network (tech-doc).
  • Endorphin fee model: measurement of computational costs and payment for execution with distribution in favor of model providers.
  • ZkMatrix (L2): a zkRollup solution to increase TPS and reduce fees (per the website).

Typical scenarios

  • Scenario 1: a developer deploys an AI contract → calls infer() → receives the model’s onchain result and uses it in the dApp’s business logic.
  • Scenario 2: an AI developer quantizes/prepares a model for MRT → publishes the model → receives a share of the endorphin fees as a “model provider”.
  • Scenario 3: a miner runs a node/miner → produces blocks → receives the block reward and part of the fees (endorphins).

Getting started (wallet/network)

  • What you need: a Cortex wallet/address; for development — the tools from the tech‑doc (including Cortex‑Remix) and the CortexTheseus client.
  • Network/chain: Cortex mainnet (PoW).
  • Fees: operations and the execution of AI contracts require endorphinLimit and endorphinPrice; the final fee = endorphinLimit * endorphinPrice (in CTXC).

Architecture/tech (non‑AI)

  • Components: the Cortex L1 network, the CortexTheseus client/node (Go), PoW mining (Cuckoo Cycle), the explorer (Cerebro), the storage layer for models/data, development tools (tech-doc).
  • Participant roles: users/dApp developers; miners; model providers (AI model providers).
  • Where logic runs: smart contracts and inference are executed onchain in the CVM; model preparation/training and quantization — offchain.
  • Dependencies: the GPU stack for nodes/mining (the docs describe the requirements), the storage network (torrentfs is mentioned in the CortexTheseus README), potential L2/zk modules (ZkMatrix per the website).

Token/economics (expanded, no forecasts)

  • Token utility: CTXC pays for computation/execution in the CVM via endorphin fees and is used in network rewards.
  • Protocol necessity drivers: without CTXC, endorphins (the execution fee) cannot be paid, and participation in the PoW reward economy is impossible.
  • Who pays fees: the senders of transactions/contract calls.
  • Where fees go: endorphins are distributed not only to miners but also to model providers (AI model providers) to motivate the publication of models (per the docs).
  • Issuance/burn: PoW: fixed block reward 7 CTXC, halving every ~4 years (per the mining docs); the total supply is stated in the docs as 299,792,458.
  • Security/incentive mechanisms: PoW consensus + fees; the “payment to model providers” model via part of the endorphins.
  • Distribution and token issuance schedule: the mining docs mention a pre‑allocation and the total mining reward; the detailed genesis/addresses were not extracted.
  • Treasury: unknown.

Governance

  • Who makes decisions: unknown (the typical process for an L1: proposals/client updates).
  • How protocol parameters change: via client/protocol updates; there are no separate governance docs in the extracted sources.
  • Where decisions are discussed/made: GitHub (CortexFoundation) and the official channels/forum on the website.
  • Keys/multisig: not applicable as a “contract admin”; for an L1, client releases and the adoption of updates by nodes/miners matter.

Security and incidents

  • Audits: the extracted Tier‑1 contains no public audit reports on the protocol/client.
  • Bug bounty: there is a public Bug Bounty Program (tech-doc), payouts in CTXC.
  • Incidents: postmortems were not extracted.
  • Threat model: the security of the PoW consensus, the correctness/determinism of inference (Synapse/MRT), GPU heterogeneity risks, risks of vulnerabilities in the client/Remix/model storage infrastructure.

Project status and timeline

  • Current status: mainnet (the CortexTheseus client, docs on mining and AI contracts).
  • Key releases/milestones: client releases are published on GitHub CortexTheseus/releases; the website describes ZkMatrix and roadmap stages.
  • Roadmap: the website lists milestones (2022–2023) and stages (consistency/scalability/privacy), as well as ZkMatrix as an L2.

Ecosystem/integrations

  • Integrations: CortexRemix (browser IDE), the CortexTheseus client, storage tooling (torrentfs is mentioned in the README), ZkMatrix (L2).
  • Partners: not extracted.
  • Where it’s actually used: AI contracts and onchain inference; ecosystem examples require a separate review of the dApps/models on the network.

AI component of the project

  • What is real AI here? Onchain inference of ML models via the CVM (infer instructions) and Synapse; models are quantized/prepared for MRT for deterministic execution.
  • What is truly onchain? The infer/inferArray calls and the execution of inference in the CVM, the storage of models/data, the calculation of endorphin fees and their distribution to miners and model providers.
  • Why is a token needed? CTXC pays the endorphin fees and is used for PoW and model provider rewards.
  • Red flags (based on the information found): AI‑washing — low (there is a technical description of CVM/Synapse/MRT and a tutorial for AI contracts); onchain‑washing — low (inference is claimed as onchain and deterministic); token‑washing — low (the token is protocol‑level: fees/rewards). What to double‑check: the availability/contents of the 2.0 whitepaper (URL in links_*; during the check the page was not extracted due to a timeout).

DecAI Fit component summary

  • DecAI Fit: 5/6 — high alignment with the Decentralized AI thesis
Component Level Score
AI reality high 3
Onchain necessity high 3
Token necessity medium 1
Evidence Tier‑1 present 3
Details

AI reality

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

  • AI_0 (decision): 0
  • Claim (one line): AI is described as the core of an already existing architecture (CVM/Synapse), not only as a plan in the roadmap.
  • Evidence: Tier‑1 — https://www.cortexlabs.ai/ ; https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/cvm.md
  • Why sufficient / why not (one line): this is a description of an approach being implemented, not an “intent”.
  • Conflict note (if any): —

AI_1 — what the AI does (1/0)

  • AI_1 (decision): 1
  • Claim (one line): the CVM executes inference of ML models onchain via the infer/inferArray instructions and the deterministic engine Synapse.
  • Evidence: Tier‑1 — https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/cvm.md ; https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/ai-contracts.md ; https://www.cortexlabs.ai/
  • Why sufficient / why not (one line): the tech docs describe the mechanism and the API for calling a model in a contract.
  • Conflict note (if any): —

AI_2 — demos/examples (1/0)

  • AI_2 (decision): 1
  • Claim (one line): the docs provide an example of an AI contract in Solidity that calls infer/inferArray.
  • Evidence: Tier‑1 — https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/ai-contracts.md
  • Why sufficient / why not (one line): there is a reproducible API example and a code skeleton of the contract.
  • Conflict note (if any): the example contains no metrics/model, but confirms the existence of the interface.

AI_3 — metrics/quality (1/0)

  • AI_3 (decision): 0
  • Metrics: —
  • Context (version/date/how measured): —
  • Evidence: Tier‑1 — https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/cvm.md
  • Why sufficient / why not (one line): the extracted fragments contain the architecture, but no public metrics of the quality/performance/accuracy of models.
  • Conflict note (if any): —

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

  • AI_4 (decision): 1
  • Claim (one line): the CVM (EVM‑compatible + infer instructions) and Synapse (deterministic inference engine) are described, as well as the MRT approach (fixed‑point).
  • Evidence: Tier‑1 — https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/cvm.md ; https://www.cortexlabs.ai/
  • Why sufficient / why not (one line): there is a technical description of the key components and their role in deterministic onchain inference.
  • Conflict note (if any): —

Onchain necessity

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

  • OC_1 (decision): 1
  • Chain: Cortex mainnet
  • Addresses/identifiers: the Cerebro explorer: https://cerebro.cortexlabs.ai/ (Remix: https://cerebro.cortexlabs.ai/remix)
  • Evidence: Tier‑1 — https://www.cortexlabs.ai/cortex-blockchain ; https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/ai-contracts.md
  • Why sufficient / why not (one line): the website and the tech docs point to the Cortex chain and the tools (Cerebro/Remix) for development.
  • Conflict note (if any): the specific chain id/addresses of system contracts were not extracted.

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

  • OC_2 (decision): 1
  • What useful logic is onchain: inference of models in the CVM, the infer/inferArray calls, the Endorphin fee model and the distribution of part of the fees to model providers.
  • Evidence: Tier‑1 — https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/cvm.md ; https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/endorphin.md
  • Why sufficient / why not (one line): this is protocol‑level useful logic onchain that goes beyond “just a token”.
  • Conflict note (if any): —

OC_3 — onchain is critical (1/0)

  • OC_3 (decision): 1
  • “Replace the blockchain” test: without onchain, the consensus verification/reproducibility of inference and the onchain economy of model provider rewards disappear.
  • Evidence: Tier‑1 — https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/cvm.md ; https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/endorphin.md
  • Why sufficient / why not (one line): Cortex directly describes the goal — to make ML inference “decentralized, immutable, transparent” and verifiable by consensus.
  • Conflict note (if any): —

Token necessity

TK_0 — no token (1/0)

  • TK_0 (decision): 0
  • Claim (one line): CTXC is the native token of Cortex, used for fees and produced by mining.
  • Evidence: Tier‑1 — https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/ctxc.md ; https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/mining.md
  • Why sufficient / why not (one line): the docs describe CTXC as the native token and the PoW issuance.
  • 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): CTXC is used to pay the execution fee (endorphins) for transactions/contracts, including AI contracts.
  • Evidence: Tier‑1 — https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/ctxc.md ; https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/endorphin.md
  • Why sufficient / why not (one line): the docs describe the fee model and the fee calculation.
  • Conflict note (if any): —

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

  • TK_2 (decision): 1
  • Claim (one line): PoW miners receive a block reward of 7 CTXC and fees; part of the endorphins goes to model providers as an incentive.
  • Evidence: Tier‑1 — https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/mining.md ; https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/endorphin.md
  • Why sufficient / why not (one line): the docs directly describe the block rewards and the distribution of fees.
  • Conflict note (if any): —

TK_3 — governance (1/0)

  • TK_3 (decision): 0
  • Claim (one line): governance mechanics of the CTXC token are not described in the extracted Tier‑1 sources.
  • Evidence: Tier‑1 — https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/ctxc.md
  • Why sufficient / why not (one line): CTXC is described as a fee/mining token; governance is not claimed.
  • Conflict note (if any): —

TK_4 — token is replaceable (1/0)

  • TK_4 (decision): 1
  • Claim (one line): as a PoW L1 with its own fee model and rewards, CTXC is the protocol currency; the token cannot be replaced without changing the protocol.
  • Evidence: Tier‑1 — https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/mining.md ; https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/endorphin.md
  • Why sufficient / why not (one line): the economic model of PoW and fees is tied to CTXC.
  • Conflict note (if any): —

Evidence (table)

Tier Source What it proves
Tier‑1 https://www.cortexlabs.ai/ Official positioning: onchain AI inference, CVM, Synapse, MRT, AI dApps.
Tier‑1 https://www.cortexlabs.ai/cortex-blockchain Description of the Cortex Blockchain and the roadmap/ZkMatrix (L2).
Tier‑1 https://github.com/CortexFoundation/tech-doc Official tech documentation on CVM/AI contracts/mining/fees/bug bounty.
Tier‑1 https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/cvm.md CVM: EVM‑compatible + infer instructions; Synapse as the deterministic inference engine; MRT.
Tier‑1 https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/ai-contracts.md The infer/inferArray interfaces and an example of an AI contract in Solidity.
Tier‑1 https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/endorphin.md The Endorphin fee model and the distribution of fees between miners and model providers.
Tier‑1 https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/mining.md PoW (Cuckoo/CuckARoo), block time ~15s, block reward 7 CTXC, halving, total supply.
Tier‑1 https://raw.githubusercontent.com/CortexFoundation/tech-doc/master/bug-bounty.md Public Bug Bounty Program (payments in CTXC).
Tier‑1 https://cortexlabs.ai/cortex_2_0_whitepaper_en Tier‑1 whitepaper link (the contents were not extracted during the check: timeout).
Tier‑2 https://www.coingecko.com/en/coins/cortex Reference profile (market/links), does not replace the tech docs.
Tier‑3 https://coinmarketcap.com/currencies/cortex/ Reference aggregator.

No comments to display

Back to top