The upcoming Mesa upgrade marks an important step forward for the Mina Protocol — improving performance, expanding zkApp capabilities, and making future upgrades easier through automation.
Before we get there, there’s still a lot to test. This post outlines how we’ll validate Mesa across multiple layers —from thorough internal testing to community-wide upgrade drills —to ensure a smooth, safe hard fork.
What Mesa Includes
Mesa will bundle several Mina Improvement Proposals (MIPs) that together improve performance and developer experience:
- Slot time reduction: from 180s to 90s for faster throughput and better UX.
- Expanded on-chain state limits: enabling richer, more flexible zkApps.
- Enhanced Event and Action Limits: enabling more expressive zkApps that can execute complex logic within a single transaction while reducing costs.
- Higher Account Update Limits: removing the need to split complex workflows across several transactions and simplifying developer UX.
- Automated hard fork mechanism: making future upgrades simpler and less manual.
- Note: this isn’t a MIP but a meaningful parameter change and an updated upgrade mechanism that is being deployed for Mesa.
These changes touch every part of the protocol, from block production and performance to zkApp behavior, which is why we’re taking the time to test thoroughly before mainnet activation.
Testing Plan Overview
Our testing strategy follows five phases, moving from internal validation to full community participation. Each phase builds on the last to eliminate risk, uncover edge cases, and strengthen confidence before the hard fork.
Phase 1: Internal Testing
We’re starting with an intensive period of internal testing — the foundation of everything that follows. The o1Labs engineering teams are running unit, integration, and end-to-end tests to validate every new Mesa feature and configuration, from slot-time reduction to on-chain state expansion and the automated upgrade system itself. These tests cover not only how components work in isolation, but how they behave together under real-world conditions.
Beyond correctness, we’re also putting Mesa through performance and load testing. This includes simulating high-traffic network scenarios to evaluate how well nodes handle faster 90-second slots, how zkApp throughput scales, and how the network behaves under sustained pressure. Benchmarking before and after Mesa will measure the improvements and uncover any bottlenecks.
Both upgrade paths — the new automated hard fork and the legacy manual process — will be thoroughly tested through a series of dry-run upgrades:
- Berkeley → Berkeley (Legacy): a manual rehearsal of the existing upgrade process to identify steps that can be streamlined or automated.
- Berkeley → Berkeley (Auto): the same process run through the new automated system to validate the infrastructure in isolation, without the new Mesa features and configs. This dry run will also allow us to prepare the runbooks for ‘Auto’ mode.
- Berkeley → Mesa (Legacy): a full manual upgrade rehearsal, including all new features, and to take us from a Berkeley network to the new Mesa network.
- Berkeley → Mesa (Auto): the final drill, mirroring what most node operators will experience for the actual hard fork.
In all of these tests, both in ‘Legacy’ and ‘Auto’ mode, the o1Labs team will be checking on various features including and not limited to: Mina Daemon, the archive node and schema update, Rosetta service compatibility and Mina OCaml Signer compatibility, o1js compatibility and Mina-Signer SDK, network consensus and block production with upgrade data extraction and recovery flows, and Mina-lightnet-docker.
Phase 2: Mesa Testnet
We’ll soon be opening the doors to the community with the Mesa Testnet (MT) — a dedicated environment to explore all new Mesa features before mainnet. This network won’t test the upgrade mechanism itself, but rather the post-upgrade Mesa state. Node operators, block producers, zkApp builders, and community members will be able to interact with the upgraded network, surface issues early, and provide feedback.
The Mesa Testnet will launch in two stages:
- An early pre-flight nightly build before code freeze. This is for developers who want to get a head start testing zkApps or node setups, and will be a very unstable network. Expect fast iteration and potential resets at this stage.
- A stable long-running testnet post–code freeze. This will be the main sandbox for node operators, block producers, and zkApp developers to experiment with the new 90-second slot times, expanded state limits, and increased action/event capacity.
This testnet serves as an important checkpoint before the community-wide upgrade rehearsals begin.
Phase 3: Mesa Trail — the Upgrade Testnet
The heart of our testing and community collaboration will be Mesa Trail, the full-scale Mesa Upgrade Testnet (MUT). This network will begin from a Berkeley-like state and execute a complete upgrade to Mesa — exactly as mainnet will. We’ve named this phase Mesa Trail, marking the journey that takes Mina from Berkeley to Mesa, and the starting point of our collective climb with builders and node operators towards the Mesa Upgrade.
Launching Mesa Trail requires several conditions to be met:
- all internal testing of Mesa features and configurations must be complete,
- load tests must show stable performance,
- on-chain voting has finished and final decisions on which MIPs are included has been made, and
- the Mesa Testnet has been spun up, tested and proven reliable for some time.
Once these are in place, we’ll release the final pre-upgrade Mina build, which includes the automated upgrade package for node operators.
To ensure we capture high-quality feedback across diverse conditions, we’re launching the Trailblazers Program — an incentivized initiative for a select group of experienced node operators chosen to represent different geographies, hardware configurations, and deployment setups. Trailblazers will help us validate both the automated upgrade process and all Mesa functionality under real network conditions. They’ll monitor block production, consensus behavior, performance, and stability after the upgrade, ensuring the network behaves as expected across every configuration.
Every insight from the Trailblazers Program benefits the entire Mina network. Findings and results from these runs will be shared publicly so that all node operators can learn from the process ahead of the mainnet upgrade.
In parallel, we’ll invite zkApp developers to join Mesa Trail to deploy and test their applications in the upgraded Mesa environment. This phase gives builders the chance to explore everything Mesa unlocks — from expanded on-chain state and higher event and action limits, to faster slot times and improved throughput. Developers can stress-test their applications, refine performance, and report any issues that surface under real-world conditions.
Together, Mesa Trail and the Trailblazers Program represent the most collaborative stage of the Mesa journey — a complete end-to-end rehearsal that combines upgrade testing, feature validation, and community participation. Once this phase runs successfully and the network operates stably in the Mesa state, we’ll be ready to announce the final date for the Mesa mainnet upgrade.
Phase 4: Ecosystem Validation
After Mesa Trail proves stable, testing will shift to the wider ecosystem — ensuring that all infrastructure and integrations around Mina can handle Mesa smoothly. We’ll coordinate directly with exchanges to validate deposits, withdrawals, and API behavior during and after upgrades; with wallets and explorers to confirm UI, API, and transaction display updates; and with infrastructure providers like RPC endpoints, indexers, and SDK maintainers to ensure all services remain compatible.
This is when ecosystem partners and tool builders can start confirming readiness, ensuring that every part of the Mina stack transitions smoothly to Mesa.
Phase 5: Devnet Hard Fork - the Final Rehearsal
Finally, we’ll perform one last end-to-end rehearsal on Devnet, which mirrors mainnet conditions without impacting live operations. This is the final confidence check before the upgrade. We’ll deploy Mesa in full and test its stability, performance, and node resilience under real workloads.
The Devnet upgrade will also serve as a final opportunity for community members and partners to test integration one last time — exchanges, wallets, explorers, zkApp developers, and node operators alike.
Our Commitment
The Mesa upgrade represents an important step in Mina’s evolution, advancing the protocol’s performance, developer experience, and upgrade infrastructure. As with every hard fork, we’re approaching it with care, precision, transparency, and necessary time to ensure the process runs smoothly and safely.
Each testing phase adds another layer of assurance. By the time Mesa reaches mainnet, every component will have been validated multiple times under real-world conditions.
For the community, there are several ways to take part:
- zkApp developers can begin preparing for the Mesa Testnet, experimenting with expanded state limits and faster slot times.
- Node operators can follow updates from the Trailblazers Program and upcoming upgrade rehearsals to understand when participation opens.
- Ecosystem partners — including wallet providers, explorers, and exchanges — will have opportunities to test integrations during the Mesa Trail. o1Labs will be in touch with these partners soon.
- And for everyone else, feedback and participation through governance discussions and MIP reviews remain vital as we finalize Mesa’s scope and timing.
We’re grateful for the continued support and collaboration of Mina’s community. Together, we’re making sure the next stage of Mina’s evolution — and every upgrade beyond it — is as reliable, inclusive, and verifiable as the protocol itself.
Keep an eye out for lots more content on Mesa as we continue the climb together.
About Mina Protocol
Mina is the world’s lightest blockchain, powered by participants. Rather than apply brute computing force, Mina uses advanced cryptography and recursive zk-SNARKs to design an entire blockchain that is about 22kb, the size of a couple of tweets. It is the first layer-1 to enable efficient implementation and easy programmability of zero knowledge smart contracts (zkApps). With its unique privacy features and ability to connect to any website, Mina is building a private gateway between the real world and crypto—and the secure, democratic future we all deserve.