Related Links: Path to Crypto AI | https://linktr.ee/0xjacobzhao

1. Project Overview:

Sentient is an open-source protocol platform dedicated to building a decentralized artificial intelligence (AI) economy. Its core goal is to establish ownership structures for AI models, provide on-chain calling mechanisms, and create composable, revenue-sharing AI Agent networks. Through the OML framework (Open, Monetizable, Loyal) and model fingerprint technology, Sentient addresses the fundamental issues in the current centralized LLM market, such as "unclear model ownership, untraceable calls, and unfair value distribution."

• Official Website: https://www.sentient.xyz

• GitHub: https://github.com/sentient-agi

• Foundation: https://sentient.foundation

• Documentation: https://docs.sentient.xyz

• Blog: https://www.sentient.xyz/blog

• X (Twitter): https://x.com/SentientAGI

The project is driven by the Sentient Foundation, a non-profit organization focused on building open-source AGI and protocol incentive mechanisms. The concept of "Loyal AI" refers to an open AI model ecosystem that serves the community, ensures fair governance, and can self-evolve in the long term.

The architecture of the Sentient Protocol consists of two core components: Blockchain System and AI Pipeline.

The AI Pipeline is the foundation for developing and training "Loyal AI" artifacts, consisting of two core processes:

• Data Curation: A community-driven data selection process for model alignment.

• Loyalty Training: Ensures that models remain consistent with the community's intent during training.

The blockchain system provides transparency and decentralized control for the protocol, ensuring AI artifacts' ownership and governance. The main modules include:

• Governance: A decision-making system controlled by a decentralized autonomous organization (DAO).

• Ownership: Ownership of AI artifacts represented via tokenization.

• Decentralized Finance (DeFi): Provides financial tools supporting open, decentralized, and fair governance and rewards.

2. Technical Architecture and Model Ownership Mechanism:

1. OML Framework

In the 2024 whitepaper "Sentient: Loyal AI" (Link to Paper), the project introduces the OML framework, which begins with model ownership and aims to build a "chain-based AI ownership protocol economy." The concept of "AI-native cryptography" is introduced, emphasizing that relying solely on code licenses and governance reputation is insufficient for the long-term development of open-source AGI. The framework aims to provide encryption-level ownership protection for open-source models.

Core Paper Explanation — "OML: Open, Monetizable, and Loyal AI":

• Open: Models must be open-source, with transparent code and data structures, supporting community reproduction, audit, and forking.

• Monetizable: Every model call should trigger a revenue stream, distributed via on-chain contracts to trainers, deployers, and validators.

• Loyal: Models belong to the contributor community, not companies, and their upgrades and governance are decided by the DAO. Model ownership must be verifiable, modifications are limited, and usage is controlled.

OML is not just a code license agreement but a mechanism that uses on-chain methods and cryptographic tools to ensure that even open models have clear ownership, economic sovereignty, and governance rights.

AI-native Cryptography： leverages the continuity, low-dimensional manifold structure, and differentiability of AI models to develop a "verifiable but non-removable" lightweight security mechanism.

Core Technologies:

• Model Fingerprinting: Embeds a unique signature into the model during training, using hidden query-response pairs.

• Ownership Verification Protocol: Uses third-party detectors to verify ownership via query-based probing.

• Permissioned Calling Mechanism: Requires permission from the model owner to authorize a call, ensuring behavior-based authorization and ownership verification.

OML Components:

Sentient currently uses the Melange method, combining fingerprinting, TEE execution, and on-chain contract revenue distribution.

Sentient Protocol and OML
The Sentient Protocol architecture supports OML and includes four main layers:

• Storage Layer: Stores model weights and fingerprint registration.

• Distribution Layer: Manages model call permissions through authorization contracts.

• Access Layer: Verifies user authorization before accessing the model.

• Incentive Layer: Routes earnings from model calls to trainers, deployers, and validators.

The protocol is compatible with Ethereum and other Layer 1 chains, and the models are registered on-chain, ensuring clear ownership and governance. The system will integrate ZK Proof (zero-knowledge proofs) for verifying model output authenticity.

2. Fingerprinting and Model Ownership Mechanism

GitHub: https://github.com/sentient-agi/oml-1.0-fingerprinting

This repository is the first implementation of Sentient's fingerprinting mechanism, providing an interface for embedding fingerprint injection and verification into the training process. Its purpose is to ensure that model ownership is verifiable, usage behavior is traceable, and to prevent unauthorized replication and commercialization. This is the specific engineering implementation of the OML (Open, Monetizable, Loyal) framework.

Functions of the Model Fingerprinting Module (OML 1.0 Fingerprinting Module):

• Behavioral Profiling: Extracts features from the model’s behavior to generate a unique fingerprint.

• Model Attribution: Determines the originating model based on the generated content.

• Fingerprint Verification: Verifies the source and consistency of the model's output.

The essence of the fingerprint mechanism is: by fine-tuning the model, embedding a set of unique "key-response" pairs, the model owner can verify whether the model belongs to them through a specific query, thus forming the model’s "cryptographic signature." In short, this acts as a "watermark" for LLMs (Large Language Models), enabling tracking and enforcement of AI ownership.

3. Enclave TEE Computing Framework

GitHub: https://github.com/sentient-agi/Sentient-Enclaves-Framework

The Sentient Enclaves Framework is an open-source framework that leverages trusted execution environments (TEE), such as AWS Nitro Enclaves, to securely deploy model inference, fine-tuning, and agent services. This framework emphasizes the "loyalty" of the model, ensuring that models respond only to authorized requests, preventing unauthorized access and use. The security features include:

• Payload Encrypted Deployment: The model, runtime configuration, and input-output are all encrypted and stored within the enclave.

• TLS-based RPC: Calls are made using bidirectional TLS with an attestation signature channel for secure communication.

• Log and Cache Isolation: Runs in transient memory, without permanent logging.

• Attestation Reports: Every call generates an output proof bound to the enclave's signature, ensuring that execution behavior is traceable and verifiable.

The TEE (Sentient Enclaves Framework) is advantageous in terms of high performance and cloud integration, making it suitable for real-time AI and sensitive data processing, but it is limited by hardware dependencies and side-channel attacks. The Sentient Enclaves Framework uses Nitro Enclaves to provide enterprise-grade privacy protection, with a relatively user-friendly development experience.

Compared to other encryption technologies, FHE provides strong privacy guarantees without hardware dependencies and quantum security, but with significant performance overhead, making it suitable for specific cryptographic computing scenarios, and is difficult to directly replace the high-performance tasks of TEE. ZK (Zero-Knowledge Proofs) performs excellently in verifiability and decentralized scenarios and can complement TEE (this module plans to integrate with zkML in the future).

4. Sentient Agent Framework

GitHub: https://github.com/sentient-agi/Sentient-Agent-Framework

The Sentient-Agent-Framework is a lightweight open-source framework focused on automating web tasks (such as search, video playback) through AI agents that control browsers. It combines natural language instructions with LLMs (e.g., OpenAI’s GPT-4o) to provide a simplified development experience (claimed to be just 3 lines of code). Its asynchronous execution, custom instructions, and multi-provider support make it suitable for rapid development and experimental applications. When combined with other Sentient AGI projects (like the TEE framework), it can be extended to security-sensitive scenarios.

Core Architecture:
This architecture supports building agents with a complete Perception–Planning–Execution–Feedback loop, which can be extended into a multi-agent collaborative, on-chain verifiable, and alignable open-source AI system.

• User Layer: Users input task goals via natural language.

• Agent Framework Layer (Sentient Agent Framework):

  • Perception: Understands input and the environment.

  • Planning: Generates action plans based on goals.

  • Execution: Executes tasks by calling tools/skills.

  • Reflection: Analyzes feedback to optimize behavior.

  • Memory: Manages short-term and long-term memory.

  • Skills/Tools: Registers external functions and plugins.

• Collaboration Layer (Multi-Agent Layer): Multiple agents collaborate, distribute tasks, and summarize results.

• External Integration Layer:

  • Blockchain Smart Contracts: Records tasks and settles incentives.

  • Agent Registry: Provides identity authentication and version control.

  • ZK Proof Module: Verifies output authenticity.

  • External Tool APIs: Includes web search, databases, file systems, etc.

• Storage Layer: Local storage + decentralized storage (e.g., IPFS).
  

When compared to traditional AI Agent Frameworks, the Sentient-Agent-Framework is more lightweight and simpler. In comparison to other Crypto AI Frameworks like Virtuals Protocol and ai16z (elizaOS), which offer diverse solutions for AI agent development, on-chain automation, or Web3 integration, Sentient-Agent-Framework is more suitable for off-chain web tasks.

5. Sentient Social Agent

GitHub: https://github.com/sentient-agi/Sentient-Social-Agent

Sentient-Social-Agent is an AI system designed to automate interactions on social platforms (such as Twitter, Discord, and Telegram). It is capable of understanding social contexts, generating content, interacting with users, and engaging in social communication through multi-agent collaboration. It utilizes social perception, content generation, and behavior planning modules to support natural conversation and content creation on platforms. This system is suitable for use cases such as brand management, virtual community management, and information dissemination. The system can also integrate with the Sentient Agent Framework.

6. Open Deep Search (Not Launched Yet)

On the Sentient website, Open Deep Search is defined as a search agent that can surpass ChatGPT and Perplexity Pro. Team member Sewoong Oh disclosed part of the plans at the EthDenver 2025 Open AGI Summit:

Open Deep Search consists of two main components: Sentient’s search functionality (which includes query rewriting, URL and document processing, etc.) and the reasoning agent. The reasoning agent utilizes open-source LLMs (such as Llama 3.1 and DeepSeek) and enhances search quality through tools like search, calculators, and self-reflection. In the Frames Benchmark, Open Deep Search outperforms other open-source models and even competes with some closed-source models. However, since its functionality is not yet launched, we are currently unable to evaluate its true capabilities.

3. Product Forms, Landing, and Plans

Currently, the main products showcased on Sentient’s official website are Sentient Chat and the open-source model Dobby LLMs:

Sentient Chat:
Sentient Chat is a decentralized AI chat platform launched by Sentient Foundation, aimed at providing a community-driven, customizable, and loyal AI experience. The platform integrates open-source large language models (such as the Dobby series) with advanced reasoning agent frameworks, supporting various tool integrations to meet diverse user needs. Key features include:

1. Open Reasoning Agents: The reasoning agents built into Sentient Chat can perform complex tasks and support the following functions:

   • Search Tool: Integrates Open Deep Search (ODS), offering real-time web search capabilities.

   • Calculator: Handles mathematical calculations and data analysis tasks.

   • Code Execution: Generates and runs Python code to perform complex logical reasoning and task execution.
     

2. Multi-agent Integration: The platform supports the integration of multiple AI agents, allowing users to choose different agents based on their needs, enhancing the platform's flexibility and functionality. This is similar to a Web3 version of POE or an open, agent-driven Perplexity alternative.

Sentient Chat is currently in the testing phase and is only accessible via invitation codes distributed through email or community events. According to official information, over 5,000 users have successfully gained access to Sentient Chat, with more than 100,000 user queries processed. Since the author is not yet a part of the test whitelist, the true capabilities of the model cannot be evaluated at this time.

Dobby LLM Model Series:
Based on Meta’s Llama series, fine-tuned on Hugging Face (Link)

1. Dobby-Unhinged Series:

   • Dobby-Unhinged-Llama-3.3-70B: Fine-tuned on Llama 3.3-70B-Instruct, emphasizing personal freedom and cryptocurrency stances, with a frank, humorous, and humanized conversational style.

   • Dobby-Mini-Unhinged-Llama-3.1-8B: The 8B parameter version, suitable for resource-constrained devices, retains the core features of the "Unhinged" series.

2. Dobby-Leashed Series:

   • Dobby-Mini-Leashed-Llama-3.1-8B: Compared to the "Unhinged" version, the tone is milder, suitable for applications requiring more stable output.

Since Dobby LLM models are fine-tuned versions of Llama 3.1 and 3.3, their primary application scenarios are believed to be in building chatbots, content generation and creation, role-playing agents, etc. Their advantages lie in flexible style generation, enhanced reasoning, and low resource requirements, making them ideal for quick deployment and flexible customization in resource-limited environments. However, compared to more powerful closed-source models like GPT-4, Dobby LLM still has limitations in handling advanced logic, cross-domain knowledge reasoning, and deep reasoning tasks.

4. Ecosystem Cooperation and Landing Scenarios

The Sentient Builder Program currently provides $1 million in funding to support developers in building AI Agents that operate within the Sentient Chat ecosystem. Developers are required to use Sentient's development kit and integrate through the Sentient Agent API to connect to the ecosystem.

At the same time, Sentient's official website lists a wide range of ecosystem partners, covering various sectors within Crypto AI. The specific list of partners includes:

• Models: Eigenlayer, Move, CrunchDAO, Bagel, KGEN

• Agents: Messari, Franklin Templeton, Kaito, MyShell, Third Web, Theoriq, Open, QNA3, Pond, Mira, Olas, Biconomy, Talus, Zettablocks, Axal, Morpheus AI, dFusion, ExponentAI, Fetch AI, Giza, JustTX, UnifAI, Questflow, QuillAI, Raiinmaker, Solo, Spectral, UOMI, PlayAI

• Data: Kaito, Vana, The Graph, Space and Time, 0g, Open, QNA3, Zettablocks, Chainbased, dFusion, Dria/First Batch, Entrova, FractionAI, Hyve DA, Irys, Masa, Mizu, OpenLedger, Raiinmaker, Sapien, Zus Network

• Verifiable AI: Nillion, Lagrange, pi2

• Blockchain: Arbitrum, Polygon, Celo

• Infrastructure: Lit Protocol, OpenGradient

As a leading project in the Crypto AI field, Sentient has the resource integration capacity to cover any prominent startup projects within the industry. However, it should be noted that "marketing-oriented" collaborations are widespread in the Crypto space, often creating a false sense of prosperity within the industry. The contribution and loyalty of Sentient's ecosystem partners to its ecosystem still require ongoing observation.

Open AGI Summit is a global conference organized by the Sentient team, focused on exploring the intersection of Artificial Intelligence (AI) and Crypto technologies. The author had the privilege of attending the summit during ETH Denver and ETHcc in 2024 and 2025. The Sentient team has demonstrated the ability to gather leading industry investors and project entrepreneurs, which remains a key highlight.

5. Team Structure and Research Background

Sentient Foundation brings together top academic experts, entrepreneurs, and engineers from the crypto industry, with a focus on building a community-driven, open-source, and verifiable AGI platform. According to the official team information (Link), the team members include:

Core Leadership (Steering Committee)

• Pramod Viswanath – Professor at Princeton University, expert in information theory and communication systems, leading Sentient's AI security and theoretical foundation development.

• Himanshu Tyagi – Professor at the Indian Institute of Science, specializing in privacy protection and decentralized learning algorithms, providing academic support for model training and privacy cooperation.

• Sandeep Nailwal – Co-founder of Polygon, responsible for blockchain strategy and global ecosystem layout, a key figure connecting the crypto community and AI architecture.

Sensys Team – A Web3-native product studio, leading user experience optimization and developer infrastructure development, driving Sentient product implementation.

Core Engineering and Development Team: Members come from well-known tech and blockchain companies, such as Meta, Coinbase, Circle, Polygon, Binance, as well as researchers from top universities including Princeton University, Washington University, and the Indian Institute of Technology. The team has extensive experience in LLM engineering, system security, computer vision, and data systems construction.

AI Research and Model Training Team: The research team covers AI/ML, NLP, computer vision, and reinforcement learning, with members who have practical experience at Google Research, Daimon Labs, Fetch.ai, and other institutions. The team composition shows that Sentient possesses both strong academic depth and implementation capabilities, as well as experience in the crypto ecosystem.

It is important to note that Sentient was initially launched with the successful background of Sandeep Nailwal, co-founder of Polygon. As an important expansion solution for the Ethereum ecosystem, Matic started with the Plasma technology, which wasn't leading but was sufficiently "cheap and fast," creating a moat for Polygon in fields like NFTs and social media. Additionally, through the acquisition of Mir Protocol and Hermez Network, and the launch of Polygon zkEVM, ZK technology was integrated into its blockchain scaling solutions.

As Sandeep Nailwal’s second entrepreneurial venture, Sentient benefits from his experience, capital, connections, and market recognition, which far exceed his previous work. It is poised to raise significant funds in 2024, even with an imperfect project concept. However, the AI field is different from crypto, and Sentient still faces external challenges such as changes in the market environment, intensified competition, and rapid technological advancements.

6. Fundraising and Token Model

Fundraising Time: Mid-2024
Amount Raised: $85 million (Seed Round)
Investing Institutions: Founders Fund, Pantera, and Framework Ventures co-led the investment. Other VC institutions include Ethereal, Robot Ventures, Symbolic Capital, Dao5, Delphi, Primitive Ventures, Nomad, Hack VC, Arrington Capital, Hypersphere, IDG, Topology, Protagonist, Folius, Sky9, Canonical Crypto, Dispersion Capital, Mirana, Foresight, Hashkey, Spartan, Republic, Frontiers Capital, etc. (Link)

$SENT Token Use (Planned):
As of now, Sentient has not yet launched an official token. Co-founder Sandeep Nailwal stated that there are currently no plans to issue a token, but this may be evaluated in the future based on community demand and project development. Therefore, any projects claiming to offer $SEN or other Sentient-related tokens should be approached with caution to avoid fraud.

According to the whitepaper, potential uses of the SENT token include:

• Mapping Agent Incentive Points to tokens.

• Proposal and voting for model version management.

• Staking to validate the authenticity of Agent outputs.

• DAO governance and dividend mechanisms, etc.

Sentient is a high-profile project that was born with significant backing. Its investor background, fundraising scale, and valuation put it ahead of most Crypto AI projects in the market. On one hand, its strong resource backing enables easier integration of industry resources, and its high fundraising allows for the recruitment of top talent and provides ample capital to support project development across industry cycles. On the other hand, the current Crypto industry tends to be disillusioned with high-valuation projects backed by VCs. Additionally, VC-backed token projects often have token prices driven by capital operations, disconnected from their fundamentals. If Sentient is unable to deliver impactful Crypto AI products and ends up issuing a token based on overvaluation, it could harm the Crypto community, which is in urgent need of trust rebuilding. How the team responds to these industry challenges is something we should continue to observe.

7. Competitive Analysis and Market Position

Most Crypto AI projects in the market focus on a single domain—such as data, models, computing, training, or inference—or develop consumer-facing applications like AI Agents. Projects positioned as AI Chains include legacy blockchain platforms transitioning into AI (such as Near and ICP), or decentralized coordination and token incentive protocols like Bittensor. Sentient’s positioning is not fully aligned with these.

On the model training side, Sentient functions more as an integration platform and maintains cooperative relationships with open-source AI models in the market. On the agent side, Sentient does share some overlapping competition with projects like Talus, Olas, or Theoriq, particularly in the domain of multi-agent systems and reasoning capabilities. However, each of these projects has its own core objectives and application scenarios, meaning there remains a degree of complementarity despite the competition.

8. Conclusion

As a decentralized artificial general intelligence (AGI) protocol platform, Sentient aims to provide clear ownership structures for AI models and enable on-chain mechanisms for model calling and value distribution, addressing the issues of unclear ownership and unfairness in the current centralized LLM market. The core framework, OML (Open, Monetizable, Loyal), uses model fingerprinting and blockchain technology to ensure ownership, transparency, and fair revenue sharing for open-source models. With the backing of Sandeep Nailwal, co-founder of Polygon, Sentient has gained support from many top VC firms and AI ecosystem partners. Despite facing uncertainties, controversies, and competition, Sentient aims to become one of the standard protocols for decentralized AI ownership, advancing the development of decentralized AGI.