The Next Leap for Pastel Network: The Matisse 2.2 Mainnet Release

The release of Matisse 2.2 marks a significant milestone — introducing extensive updates aimed at enhancing network stability, refining performance, and expanding the functionalities within the ecosystem. This update is particularly notable for its focus on improving the infrastructure for AI and LLM applications across the Web3 space.

Let’s delve into how this update redefines & impacts the Pastel ecosystem.

Important Notes:

  • The latest Release Version 2.2 release activated at the Pastel Network Block Height of 683,000, or approx. May 17, 2024 at 11:00:00 AM UTC. After this Block Height, prior client versions are no longer be supported.
  • Upgrade instructions can be found here and latest binaries can be downloaded directly here.

Key Highlights of Matisse 2.2:

  • AI & LLM Services via the Inference Protocol (Beta): Matisse 2.2 lays the foundation of new ticket standards and GoNode services to power the Inference Protocol — a suite of powerful and cutting-edge AI &LLM. Powered by Pastel’s Supernodes, the Beta release of Inference will be going live on Mainnet shortly and will significantly boost the network’s utility for advanced data analysis and AI-driven tasks.
  • Enhanced Network Stability: The resolution of the duplicate-ticket issue is a major step in ensuring smoother chain transitions, robust network operations, improved performance and higher database integrity.
  • Supernode Enhancements: Updates to the mining eligibility conditions for Supernodes will optimize their functionality and ensure consistent participation in block mining, thereby bolstering the network’s efficiency and reliability.
  • API Improvements: The release features several API enhancements including improved block validation and new functionalities for services like block navigation or contract ticketing. These changes enhance developer interaction, paving the way for innovative applications.
  • Efficient Storage Fee Management: Rectifying parsing errors in the storage-fee setting RPC API ensures accurate fee adjustments, contributing to reliable cost management and predictable outlays.
  • Quality Assurance and Testing: The extensive Python test suite enhancements in this release underscore Pastel’s commitment to reliability and robust performance, ensuring that new features integrate seamlessly into the existing framework.
  • Additional Updates: The latest release bring improvements in maintenance, security, data access, navigation, and interaction, enhancing platform compatibility and system responsiveness. Security is fortified with stricter block validation and fewer unnecessary log messages. Enhanced RPC APIs now provide better insights into transaction behavior and facilitate easier blockchain navigation and interaction, moving towards a more efficient and user-friendly blockchain ecosystem.

Inference Protocol is Here

The Inference Protocol is a groundbreaking feature of the Matisse 2.2 update, laying the groundwork for new ticket-standards and GoNode services essential for powering the future of AI and LLM services across a fully decentralized environment. Set for a full release in the near future, Inference will be powered by the current Supernode network, revolutionizing how decentralized services are delivered.

Inference unlocks powerful AI-driven functionalities across localized LLMs for content generation and dynamic analysis. It empowers Supernodes to offer innovative services like AI-powered content verification via the Sense protocol, which ensures NFT authenticity. Looking ahead, Inference will facilitate hosting and execution of ML models, customizable AI services through smart contracts, and an AI marketplace to share and access a wide range of AI capabilities.

At launch, Inference activates a suite of API endpoints that streamline communication among Supernodes. This network synergy allows developers and users to purchase API credit packs using PSL, which can be used to cover the costs associated with Inference requests. The service integrates renowned AI and LLM models from top providers like OpenAI, Anthropic, Groq, and Mistral, as well as locally hosted options through the innovative Swiss Army Llama package. The Beta will be delivered via a user-friendly web-application that is can also be accessed via Pastel Wallet.

Additionally, Inference’s robust infrastructure features key components that enhance network interaction. The Async RPC client facilitates seamless interaction with the network’s RPC interface, while enhanced messaging functions allow Supernodes to exchange information securely using PastelIDs. Message broadcasting and a local database optimize message storage and retrieval, ensuring efficient communication and providing detailed insights into network performance and data analytics.

Stay tuned for more information!

As Matisse 2.2 goes live, we encourage our community to join us in this journey by upgrading their systems by blockHeight 683,000, ensuring seamless integration with these enhancements.

We’re committed to transparency and community involvement. The full stable release source code for Matisse 2.2 is available here, and we encourage developers to explore and contribute to our growing ecosystem.

Interested in our Grant’s Program? Apply here!

Looking Ahead

Matisse 2.2 represents a significant step forward in the evolution of Pastel’s Mainnet, but the journey continues. Pastel remains dedicated to driving innovation, fostering collaboration, and continuously improving the ecosystem.

Inference will be deployed to Mainnet shortly after this release. we are excited to offer the Pastel ecosystem with a comprehensive suite of AI/ML services that will support Web3 for years to come.

About Pastel Network

Pastel Network is a fully decentralized, interoperable layer-1 blockchain serving as the preeminent protocol standard for non-fungible tokens (“NFTs”) and Web3 technology.

Pastel powers layer-1 & layer-2 blockchains, decentralized applications (”dApps”), or third-party enterprises with next-generation infrastructure including NFT copymint protection, permanent data storage, and more. Lightweight protocols delivered via Web3 open APIs like Sense and Cascade can be easily integrated across existing ecosystems. A wide range of applications can be built directly on Pastel, enabling developers to enjoy the NFT primitives, composability, plug-and-play modules, and security of the broader network.

Pastel is managed by world-class developers, cryptographers, and technologists, supported alongside an experienced and extensive network of marketers, influencers, and third-party agencies. Pastel is backed by key stakeholders including Innovating Capital, a prominent venture fund.

For more information on Pastel Network, visit

Website | Twitter | Discord | Telegram | Github | Reddit | Instagram

Scroll to Top