This week, the most interesting topic was not necessarily token prices.
Nor was it another miracle narrative.
What stood out instead is that several infrastructure projects are continuing to move forward on very concrete topics: payments, RPC, deployment, tokenization, compute, and AI.
Cosmos is working on payment rails and tokenomics. Lava is pushing RPC resilience and infrastructure for tokenization. Flux is simplifying application deployment on a distributed network. Akash is moving deeper into AI compute and developer workflows.
Taken separately, these topics may seem technical.
But together, they show an interesting trend: over the past few weeks, and even months, DePIN and decentralized infrastructure have gradually been regaining space in the crypto narrative.
Not necessarily with a lot of noise.
But with use cases that are easier to understand.
And that may be exactly what the market needed.
Cosmos Hub: Restoring Cadence and Bringing Back Concrete Updates
This week, the Hub Unit published its sixth weekly recap for the Cosmos Hub. The goal of this format is to summarize important announcements, validator calls, community events, and ecosystem updates every Thursday.
Three topics stood out this week:
- the launch of the Mad Easy on Cosmos hackathon;
- progress on the tokenomics research led with Gauntlet;
- an update on the Injective USDC integration.
What matters here is not only the content of these announcements. It is also the method.
After months of difficult debates around the role of the Cosmos Hub, the future of ATOM, EVM, ICL, and the Hub’s actual place in the broader ecosystem, this weekly cadence brings at least one valuable thing: clarity.
It is not a definitive answer to all the big questions.
But it is an attempt to bring back follow-up, rhythm, and concrete updates.
And Cosmos probably needed that.
Mad Easy on Cosmos: Bringing Builders Back to the Hub
The Mad Easy on Cosmos hackathon, organized with the Mad Scientists, kicked off this week.
It is a one-week sprint focused on AI, on-chain games, and incentive experiments. Participants are invited to build prototypes around loot mechanics, gacha systems, bonding curves, social coordination, or risk/reward simulations.
The important condition: at least one part of the project must be connected to the Cosmos Hub.
That detail matters.
The Hub cannot remain only a staking asset, a governance topic, or a historical symbol of the Cosmos ecosystem. It needs to become a place where people build again.
This hackathon will not transform the Hub in one week.
But it sends a useful signal: bringing builders, prototypes, and experiments back to the center of attention.
And for an ecosystem that has sometimes spent too long debating its own identity, building may still be the best way to find direction again.
ATOM Tokenomics: Less Intuition, More Data
Another important update concerns the tokenomics research conducted with Gauntlet, which is now approaching the end of its first phase.
The goal is to better understand how ATOM moves, why it moves, when it is sold, and by which types of actors. The work relies on wallet-level analysis, holder cohorts, and sell-pressure attribution.
This is probably one of the most important workstreams for ATOM.
For a long time, many discussions around the token have been based on impressions: too much inflation, too much sell pressure, not enough value capture, not enough utility.
These questions are legitimate.
But if the Hub wants to seriously change its economic model, it needs to rely on solid data.
Otherwise, the risk is simple: replacing one assumption with another.
The Hub Unit indicated that several analyses still need to be completed before the final Phase 1 report, including purchase attribution, cohort reactions to past events, and a comparison with other proof-of-stake networks.
It is less spectacular than a major marketing announcement.
But it is probably more useful.
Injective USDC: Moving Forward Without Creating Panic
The USDC topic is also progressing.
The integration via Injective is still in technical design, but the message remains clear: there is no urgency to migrate. Noble USDC continues to work as it does today, and no deadline is being imposed on teams currently using it.
This caution matters.
In Cosmos, asset, liquidity, or standard migrations can quickly become confusing. Reminding teams that there is no need to panic helps avoid a rushed transition.
In the medium term, this topic remains strategic.
USDC, liquidity, and payment rails are not minor details. They are essential if Cosmos wants to become relevant again for broader financial use cases.
And this connects directly to another important topic this week: the possible role of Cosmos in modern payment networks.
Cosmos and Payments: Interoperability as an Institutional Argument
Cosmos also published an article on how modern payment networks can connect to blockchain infrastructure.
The message is interesting because it avoids the usual trap of saying that “blockchain will replace the entire banking system.”
The approach is more realistic: today’s payment systems are regulated, deeply integrated into institutional workflows, and built around highly controlled messaging and settlement flows. The blockchain opportunity is therefore not necessarily to replace everything, but to add settlement layers that are more transparent, programmable, and interoperable.
This is where Cosmos may have a real role to play.
Its architecture allows specialized, sovereign blockchains to be built for specific needs: payments, tokenized assets, interbank settlement, stablecoins, compliance, and dedicated governance.
Projects such as Stable and Arc are cited as examples of payment-focused networks. Circle also presents Arc as blockchain infrastructure designed for an internet-native economy, extending beyond payments into capital formation, contracts, and economic coordination.
This is not mass adoption yet.
But it is a more credible direction: building rails that can connect with institutions, rather than pretending that institutions are simply going to disappear.
Lava: When RPC Becomes Critical Infrastructure
This week, Lava also reminded the ecosystem of something essential: access to a blockchain is almost as important as the blockchain itself.
When a wallet, dApp, or multisig depends on a single RPC provider, any failure from that provider can break the user experience. Users may believe their wallet is broken, their balance has disappeared, or the network is down, while the issue may simply come from the access layer.
This is not theoretical.
Past incidents have already shown that widely used Web3 services can be affected by centralized cloud or RPC dependencies. CryptoSlate noted, for example, that an AWS outage disrupted MetaMask balance displays and slowed some operations on Base, exposing how many interfaces depend on centralized access points.
This is precisely the problem Lava aims to reduce.
Lava routes requests through a network of RPC providers, with the goal of limiting downtime and directing traffic toward the best-performing available nodes. Its documentation highlights more than 140 billion relays processed, 99.999% reported availability, and over 30 RPC providers.
For the end user, this infrastructure is invisible when everything works.
But when it breaks, everyone notices.
And that is when we understand that decentralization is not only about validators or tokens. It is also about APIs, RPCs, dashboards, wallets, and all the intermediate layers that allow applications to function.
Lava and Tokenization: RWA Will Need Strong Rails
Lava was also connected this week to a more institutional topic: real-world asset tokenization.
Bretagne Holding Limited has reportedly selected Lava to support a Tokenization Sandbox related to Alba Bay, a large real estate development planned in the Dominican Republic. According to The Defiant, this is a preliminary agreement to design a tokenization sandbox for a residential development of more than 40,000 units, spanning nearly 40 million square meters.
Caution is necessary here.
A sandbox is not final adoption. It is not proof that the entire project will be tokenized tomorrow morning.
But the signal is interesting.
Real-world asset tokenization does not only depend on smart contracts or polished presentations. It depends on infrastructure that can last: connectivity, availability, reliable data, access to networks, operational security, and integration with other platforms.
This is where actors like Lava are trying to position themselves.
Not as “the RWA project” itself, but as a technical layer that can help such projects operate under better conditions.
And this confirms a broader trend: RWA will not only be a finance story.
It will also be an infrastructure story.
Flux Orbit: Making Decentralized Deployment Simpler
On the Flux side, the announcement around Orbit / Deploy with Git moves in a complementary direction.
The idea is simple: allow developers to connect a Git repository and deploy an application on FluxCloud without needing to master Docker or handle server configuration manually.
Flux documentation states that Deploy with Git supports many frameworks and languages, including React, Next.js, Vue, Django, Rails, Laravel, Node.js, Python, Go, Rust, Java, .NET, and PHP. It also highlights deployment from GitHub, automatic updates from Git, and automatic rollback in case of a bad deployment.
This matters.
Decentralized infrastructure often suffers from the same problem: it is powerful, but too complex for many users.
If a developer needs to understand Docker, ports, servers, images, monitoring, and network constraints before even deploying a simple application, many will give up before they start.
Orbit is therefore moving in the right direction: hiding part of the complexity to make Flux infrastructure more accessible.
And this is probably how decentralized networks can gain adoption.
Not only by explaining that they are more sovereign or more resilient.
But by becoming easier to use.
AkashML: AI Also Becomes an Infrastructure Topic
The final interesting signal comes from AkashML, which now provides documentation for using Claude Code with AkashML.
Concretely, AkashML exposes an endpoint compatible with Anthropic’s Messages API. This allows Claude Code to send requests to AkashML through a few environment variables.
The implication is important.
Claude Code is increasingly used by developers to analyze repositories, modify code, automate certain tasks, or speed up development. By allowing this type of workflow to be routed through AkashML, Akash is positioning itself in a very concrete area: AI usage inside developer workflows.
The documentation includes a mapping between the tiers used by Claude Code and different models available through AkashML, such as DeepSeek, Kimi, and Qwen.
This is not only a technical integration.
It is a broader signal: decentralized infrastructure is no longer only trying to host Web3 applications. It is also beginning to connect with existing AI workflows.
And this probably deserves a full article of its own.
Because behind AkashML, open models, TAO, Kimi, Qwen, DeepSeek, and recent debates around access to certain proprietary models, a much larger question appears:
who will control access to artificial intelligence tomorrow?
We will come back to this.
A Broader Trend: DePIN Is Returning to the Narrative
What is interesting is that these announcements are not happening in isolation.
Over the past few weeks, and even months, there has been renewed interest in decentralized infrastructure projects: compute, storage, RPC, deployment, AI, physical networks, and distributed technical services.
In other words, DePIN seems to be gradually regaining a place in the crypto narrative.
Not necessarily as a wave of market euphoria.
More as an answer to very concrete problems.
When applications depend on a single RPC provider, Lava becomes relevant.
When deploying on decentralized infrastructure remains too complex, Flux Orbit becomes interesting.
When access to AI compute and models becomes a strategic issue, Akash naturally comes back into the discussion.
And when Cosmos talks about payments, institutional rails, and interoperability, we find the same idea again: crypto needs to provide useful infrastructure, not only promises.
This may be where DePIN is starting to stand out.
It is not only selling a narrative. It is trying to answer a simple question: who runs the services that Web3, AI, and tomorrow’s applications will need?
Conclusion: Less Noise, More Things That Work
In the end, this weekly shows one thing above all: infrastructure projects are becoming hard to ignore again.
They are not always the most spectacular announcements. But they often say the most about the real state of the market.
Users want services that work.
Developers want simpler tools.
Institutions want reliable rails.
And decentralized networks that can answer these needs may be the ones that stand out in the coming months.
Cosmos, Lava, Flux, and Akash are not telling exactly the same story.
But they all point in a similar direction: a crypto ecosystem less obsessed with promises, and more focused on real infrastructure.
It may not be the loudest narrative.
But it is probably one worth watching closely.
Cosmos, Lava, Flux, Akash: Infrastructure Is Regaining Ground
This week, the most interesting topic was not necessarily token prices.
Nor was it another miracle narrative.
What stood out instead is that several infrastructure projects are continuing to move forward on very concrete topics: payments, RPC, deployment, tokenization, compute, and AI.
Cosmos is working on payment rails and tokenomics. Lava is pushing RPC resilience and infrastructure for tokenization. Flux is simplifying application deployment on a distributed network. Akash is moving deeper into AI compute and developer workflows.
Taken separately, these topics may seem technical.
But together, they show an interesting trend: over the past few weeks, and even months, DePIN and decentralized infrastructure have gradually been regaining space in the crypto narrative.
Not necessarily with a lot of noise.
But with use cases that are easier to understand.
And that may be exactly what the market needed.
Cosmos Hub: Restoring Cadence and Bringing Back Concrete Updates
This week, the Hub Unit published its sixth weekly recap for the Cosmos Hub. The goal of this format is to summarize important announcements, validator calls, community events, and ecosystem updates every Thursday.
Three topics stood out this week:
- the launch of the Mad Easy on Cosmos hackathon;
- progress on the tokenomics research led with Gauntlet;
- an update on the Injective USDC integration.
What matters here is not only the content of these announcements. It is also the method.
After months of difficult debates around the role of the Cosmos Hub, the future of ATOM, EVM, ICL, and the Hub’s actual place in the broader ecosystem, this weekly cadence brings at least one valuable thing: clarity.
It is not a definitive answer to all the big questions.
But it is an attempt to bring back follow-up, rhythm, and concrete updates.
And Cosmos probably needed that.
Mad Easy on Cosmos: Bringing Builders Back to the Hub
The Mad Easy on Cosmos hackathon, organized with the Mad Scientists, kicked off this week.
It is a one-week sprint focused on AI, on-chain games, and incentive experiments. Participants are invited to build prototypes around loot mechanics, gacha systems, bonding curves, social coordination, or risk/reward simulations.
The important condition: at least one part of the project must be connected to the Cosmos Hub.
That detail matters.
The Hub cannot remain only a staking asset, a governance topic, or a historical symbol of the Cosmos ecosystem. It needs to become a place where people build again.
This hackathon will not transform the Hub in one week.
But it sends a useful signal: bringing builders, prototypes, and experiments back to the center of attention.
And for an ecosystem that has sometimes spent too long debating its own identity, building may still be the best way to find direction again.
ATOM Tokenomics: Less Intuition, More Data
Another important update concerns the tokenomics research conducted with Gauntlet, which is now approaching the end of its first phase.
The goal is to better understand how ATOM moves, why it moves, when it is sold, and by which types of actors. The work relies on wallet-level analysis, holder cohorts, and sell-pressure attribution.
This is probably one of the most important workstreams for ATOM.
For a long time, many discussions around the token have been based on impressions: too much inflation, too much sell pressure, not enough value capture, not enough utility.
These questions are legitimate.
But if the Hub wants to seriously change its economic model, it needs to rely on solid data.
Otherwise, the risk is simple: replacing one assumption with another.
The Hub Unit indicated that several analyses still need to be completed before the final Phase 1 report, including purchase attribution, cohort reactions to past events, and a comparison with other proof-of-stake networks.
It is less spectacular than a major marketing announcement.
But it is probably more useful.
Injective USDC: Moving Forward Without Creating Panic
The USDC topic is also progressing.
The integration via Injective is still in technical design, but the message remains clear: there is no urgency to migrate. Noble USDC continues to work as it does today, and no deadline is being imposed on teams currently using it.
This caution matters.
In Cosmos, asset, liquidity, or standard migrations can quickly become confusing. Reminding teams that there is no need to panic helps avoid a rushed transition.
In the medium term, this topic remains strategic.
USDC, liquidity, and payment rails are not minor details. They are essential if Cosmos wants to become relevant again for broader financial use cases.
And this connects directly to another important topic this week: the possible role of Cosmos in modern payment networks.
Cosmos and Payments: Interoperability as an Institutional Argument
Cosmos also published an article on how modern payment networks can connect to blockchain infrastructure.
The message is interesting because it avoids the usual trap of saying that “blockchain will replace the entire banking system.”
The approach is more realistic: today’s payment systems are regulated, deeply integrated into institutional workflows, and built around highly controlled messaging and settlement flows. The blockchain opportunity is therefore not necessarily to replace everything, but to add settlement layers that are more transparent, programmable, and interoperable.
This is where Cosmos may have a real role to play.
Its architecture allows specialized, sovereign blockchains to be built for specific needs: payments, tokenized assets, interbank settlement, stablecoins, compliance, and dedicated governance.
Projects such as Stable and Arc are cited as examples of payment-focused networks. Circle also presents Arc as blockchain infrastructure designed for an internet-native economy, extending beyond payments into capital formation, contracts, and economic coordination.
This is not mass adoption yet.
But it is a more credible direction: building rails that can connect with institutions, rather than pretending that institutions are simply going to disappear.
Lava: When RPC Becomes Critical Infrastructure
This week, Lava also reminded the ecosystem of something essential: access to a blockchain is almost as important as the blockchain itself.
When a wallet, dApp, or multisig depends on a single RPC provider, any failure from that provider can break the user experience. Users may believe their wallet is broken, their balance has disappeared, or the network is down, while the issue may simply come from the access layer.
This is not theoretical.
Past incidents have already shown that widely used Web3 services can be affected by centralized cloud or RPC dependencies. CryptoSlate noted, for example, that an AWS outage disrupted MetaMask balance displays and slowed some operations on Base, exposing how many interfaces depend on centralized access points.
This is precisely the problem Lava aims to reduce.
Lava routes requests through a network of RPC providers, with the goal of limiting downtime and directing traffic toward the best-performing available nodes. Its documentation highlights more than 140 billion relays processed, 99.999% reported availability, and over 30 RPC providers.
For the end user, this infrastructure is invisible when everything works.
But when it breaks, everyone notices.
And that is when we understand that decentralization is not only about validators or tokens. It is also about APIs, RPCs, dashboards, wallets, and all the intermediate layers that allow applications to function.
Lava and Tokenization: RWA Will Need Strong Rails
Lava was also connected this week to a more institutional topic: real-world asset tokenization.
Bretagne Holding Limited has reportedly selected Lava to support a Tokenization Sandbox related to Alba Bay, a large real estate development planned in the Dominican Republic. According to The Defiant, this is a preliminary agreement to design a tokenization sandbox for a residential development of more than 40,000 units, spanning nearly 40 million square meters.
Caution is necessary here.
A sandbox is not final adoption. It is not proof that the entire project will be tokenized tomorrow morning.
But the signal is interesting.
Real-world asset tokenization does not only depend on smart contracts or polished presentations. It depends on infrastructure that can last: connectivity, availability, reliable data, access to networks, operational security, and integration with other platforms.
This is where actors like Lava are trying to position themselves.
Not as “the RWA project” itself, but as a technical layer that can help such projects operate under better conditions.
And this confirms a broader trend: RWA will not only be a finance story.
It will also be an infrastructure story.
Flux Orbit: Making Decentralized Deployment Simpler
On the Flux side, the announcement around Orbit / Deploy with Git moves in a complementary direction.
The idea is simple: allow developers to connect a Git repository and deploy an application on FluxCloud without needing to master Docker or handle server configuration manually.
Flux documentation states that Deploy with Git supports many frameworks and languages, including React, Next.js, Vue, Django, Rails, Laravel, Node.js, Python, Go, Rust, Java, .NET, and PHP. It also highlights deployment from GitHub, automatic updates from Git, and automatic rollback in case of a bad deployment.
This matters.
Decentralized infrastructure often suffers from the same problem: it is powerful, but too complex for many users.
If a developer needs to understand Docker, ports, servers, images, monitoring, and network constraints before even deploying a simple application, many will give up before they start.
Orbit is therefore moving in the right direction: hiding part of the complexity to make Flux infrastructure more accessible.
And this is probably how decentralized networks can gain adoption.
Not only by explaining that they are more sovereign or more resilient.
But by becoming easier to use.
AkashML: AI Also Becomes an Infrastructure Topic
The final interesting signal comes from AkashML, which now provides documentation for using Claude Code with AkashML.
Concretely, AkashML exposes an endpoint compatible with Anthropic’s Messages API. This allows Claude Code to send requests to AkashML through a few environment variables.
The implication is important.
Claude Code is increasingly used by developers to analyze repositories, modify code, automate certain tasks, or speed up development. By allowing this type of workflow to be routed through AkashML, Akash is positioning itself in a very concrete area: AI usage inside developer workflows.
The documentation includes a mapping between the tiers used by Claude Code and different models available through AkashML, such as DeepSeek, Kimi, and Qwen.
This is not only a technical integration.
It is a broader signal: decentralized infrastructure is no longer only trying to host Web3 applications. It is also beginning to connect with existing AI workflows.
And this probably deserves a full article of its own.
Because behind AkashML, open models, TAO, Kimi, Qwen, DeepSeek, and recent debates around access to certain proprietary models, a much larger question appears:
who will control access to artificial intelligence tomorrow?
We will come back to this.
A Broader Trend: DePIN Is Returning to the Narrative
What is interesting is that these announcements are not happening in isolation.
Over the past few weeks, and even months, there has been renewed interest in decentralized infrastructure projects: compute, storage, RPC, deployment, AI, physical networks, and distributed technical services.
In other words, DePIN seems to be gradually regaining a place in the crypto narrative.
Not necessarily as a wave of market euphoria.
More as an answer to very concrete problems.
When applications depend on a single RPC provider, Lava becomes relevant.
When deploying on decentralized infrastructure remains too complex, Flux Orbit becomes interesting.
When access to AI compute and models becomes a strategic issue, Akash naturally comes back into the discussion.
And when Cosmos talks about payments, institutional rails, and interoperability, we find the same idea again: crypto needs to provide useful infrastructure, not only promises.
This may be where DePIN is starting to stand out.
It is not only selling a narrative. It is trying to answer a simple question: who runs the services that Web3, AI, and tomorrow’s applications will need?
Conclusion: Less Noise, More Things That Work
In the end, this weekly shows one thing above all: infrastructure projects are becoming hard to ignore again.
They are not always the most spectacular announcements. But they often say the most about the real state of the market.
Users want services that work.
Developers want simpler tools.
Institutions want reliable rails.
And decentralized networks that can answer these needs may be the ones that stand out in the coming months.
Cosmos, Lava, Flux, and Akash are not telling exactly the same story.
But they all point in a similar direction: a crypto ecosystem less obsessed with promises, and more focused on real infrastructure.
It may not be the loudest narrative.
But it is probably one worth watching closely.
