State of Akash Q1 2025

Key Insights

• Lease revenue surpassed $1 million (38% QoQ), driven by longer and higher-value GPU deployments.
• Average GPU capacity rose 55% QoQ to 897 units, while usage increased 54% to 553 units.
• The number of active providers increased 4% QoQ to 69, aided by the launch of the Akash Provider Console in February.
• Eight governance proposals passed in Q1 2025, including reductions to AKT inflation, an increase in the community pool tax, and funding for analytics, market liquidity, and events.
• Akash introduced autonomous AI agent deployments in Q1 2025, supported by tools from Levangie Labs and ElizaOS, enabling agents to provision compute independently on the network.

Primer

Akash (AKT) is a decentralized cloud computing marketplace that facilitates the buying and selling of compute resources. It is an open-source, permissionless protocol that provides an alternative to today’s centralized cloud services (i.e., AWS, Azure, and Google Cloud). Akash aims to leverage underutilized server capacity, which can range from 5% to over 30%. Akash is a Tendermint-based, Layer-1 network built using the Cosmos SDK. Marketplace activity (requests, bids, lease details, etc.) is stored onchain and payments are settled with Akash’s native token, AKT.

The Akash marketplace functions via a reverse auction, giving users the ability to name a price and describe the resources they want for deployments. Akash’s decentralized network of compute providers runs its open-source software and competes to provide resources, often at a fraction of the cost of big cloud providers. Specifically, Akash hosts containers where users can run any cloud-native application (e.g., AI workloads, gaming servers, blockchain nodes, and websites). Akash offers extensive cloud management services like Kubernetes, which can be used for hosting and managing containers. Additionally, Akash supports decentralized AI applications such as Venice.ai, AkashChat, and AkashGen, reflecting its role in enabling AI infrastructure. For a full primer on Akash, refer to our Initiation of Coverage report.

Website / X / Discord

Key Metrics

Usage and Provider Analysis

Leases and Revenue

Akash’s marketplace uses a reverse auction, in which users propose a bid that describes the resources they’d like to use for a deployment. When accepted, a lease is opened onchain managing the activity of this relationship.

New leases on Akash Network represent agreements between users and providers for renting computational resources. New leases on Akash declined 24% QoQ, from 61,000 in Q4 2024 to 46,000 in Q1 2025. The decrease follows a spike in lease volume during the previous quarter, which was driven by ecosystem events such as Akash’s integration with Passage (which led to GPU utilization peaking at 85% during Cosmoverse), the decentralized training of the INTELLECT-1 model, and the launch of AkashGen. These high-engagement activities led to an unusually high baseline in Q4. The Q1 drop reflects a normalization of activity levels following that burst of demand.

In Q1, activity remained elevated but was comparatively lower in absolute terms. While new integrations with projects like Envision Labs and NodeShift supported GPU-intensive workloads, these use cases are generally characterized by longer-running deployments rather than frequent new lease creation. Similarly, the introduction of AI agents and the launch of the AkashChat plugin for ElizaOS expanded autonomous infrastructure access, but the impact on lease count was less immediate. These developments likely contributed more to sustained GPU utilization than to new lease growth.

Active leases on Akash represent ongoing agreements where users utilize and manage computational resources on the network. Leases remain active as long as they are in use and cease once closed. While the number of new leases decreased by 24% QoQ, the average number of daily active leases increased by 6%, from 821 to 868 This suggests a shift toward longer-lasting workloads, where fewer new deployments were created, but existing leases remained active for longer periods. Enterprise-oriented integrations, such as NodeShift’s support for persistent VPC environments and Envision Labs' GPU inference tasks, likely contributed to this trend by increasing the duration and stability of deployments.

Daily lease revenue (USD) from lease activity grew by 41% QoQ, from $8,100 to $11,300 per day, while total quarterly lease revenue (USD) rose from $741,000 to $1 million, a 38% increase QoQ. The increase in revenue despite fewer new leases points to a higher average value per lease. This aligns with Akash’s growing emphasis on high-density GPU workloads, including support for NVIDIA H100, A100, and H200 instances. These types of deployments typically command higher prices and run longer durations, reflecting greater infrastructure demand from AI and machine learning applications.

Revenue from network transaction fees: which includes base fees, provider take rates, and any other transaction-level network costs, also rose 37% QoQ in USD terms. Notably, when denominated in AKT, the increase was more significant: from 2,700 AKT to 6,300 AKT, a 130% QoQ increase. This discrepancy is attributed to the 59% decline in the AKT token price over the same period. The rise in AKT-denominated network transaction fees indicates that demand for network services grew substantially, helping offset the effects of market-driven token depreciation.

Resource Compute

Akash’s average infrastructure usage continued to grow in Q1 2025:

• GPU Usage: Rose from 359 to 553 (+54%), supported by an increase in GPU capacity from 577 to 897 (+55%). This reflects growing demand for compute-heavy deployments, matched by sufficient supply.
• CPU Usage: Increased from 4,400 to 4,800 (+10%), while capacity rose from 20,900 to 25,600 (+23%), showing steady demand alongside consistent provisioning.
• RAM Usage: Grew from 13.6 TB to 20.5 TB (+51%), with capacity rising from 117 TB to 164 TB (+41%).
• Storage Usage: Climbed from 36.3 TB to 57.8 TB (+59%), with capacity increasing from 869 TB to 1,350 TB (+55%).

Across all major resources, both usage and capacity expanded, indicating continued demand growth and network readiness to support it.

GPU Compute

Akash’s average GPU capacity increased from 577 units to 897 units, a 55% QoQ increase. GPU usage rose from 359 to 553 units over the same period, also up 54% QoQ. The growth in both capacity and utilization indicates that additional supply was absorbed by demand, reflecting sustained usage for high-performance compute tasks.

Active Providers

Akash’s permissionless network allows resource providers to join from anywhere in the world, creating a geographically diverse and resilient infrastructure. This global distribution enhances the network’s ability to withstand regional disruptions, such as natural disasters, power outages, or political instability. It also improves performance by enabling tasks to be processed closer to end users, reducing latency and optimizing data transfer efficiency.

In Q1 2025, the number of active providers increased from 67 to 69, a 4% QoQ increase. This growth is likely due to early impact of onboarding improvements such as the launch of the Akash Provider Console in February 2025. This tool lowered technical barriers for new entrants, particularly non-crypto-native data center operators.

Token Analysis

Market Cap

AKT's circulating market capitalization declined 59% QoQ, falling from $692 million to $287 million. This drop tracked closely with the token’s price, which also fell 59% QoQ, from $2.79 to $1.16.

Qualitative Analysis

Partnerships and Developments

AI Agent Infrastructure Expansion: Akash entered a new phase of development in Q1 2025 with the introduction of autonomous AI agent deployment. On Jan 7, the network enabled AI agents to independently provision compute on the Akash Supercloud, marking the beginning of what the team calls the “Agentic Age.” This capability was enhanced by Levangie Labs, which released a framework allowing agents to control local environments and deploy to Akash without human intervention. This was followed by the launch of an Akash plugin for ElizaOS, developed by AIFlow_ML in collaboration with ai16zDAO, enabling permissionless AI agent workflows across decentralized infrastructure.

Integration with Witness Chain: InfinityWatch: In Jan 2025, Witness Chain introduced InfinityWatch, a deployment of Watchtower nodes on the Akash Network. These nodes perform delay-based internet measurements and cryptographic validation of real-world attributes, like location and connectivity. This extends the Proof of Location (PoL) paradigm and supports validation frameworks for decentralized AI agents operating on Akash, reinforcing the network’s real-world attribution capabilities.

Envision Labs Integration: Envision Labs, a decentralized AI (DeAI) platform for generative content creation, integrated with Akash to support image and video model inference in March 2025. The platform, which spans custom avatars and IP-verified generative assets, reported a 30% reduction in GPU costs after adopting Akash. During the integration period, Envision trained over 35 models and generated more than 100,000 AI-created images.

NodeShift Enterprise Deployment: NodeShift, a streamlined AI model deployment platform, integrated with Akash to power enterprise-level LLM training and inference. The collaboration allows clients to access high-density NVIDIA GPUs (including H100s, A100s, and H200s) via a familiar interface that abstracts away blockchain complexities. The deployment also validated Akash’s enterprise readiness by supporting virtual private cloud (VPC) setups, attached storage volumes, and secure multi-tenancy workloads.

Launch of Akash Provider Console: On Feb 18, Akash launched the Provider Console to simplify onboarding for new compute providers, including traditional data center operators. The interface automates Kubernetes installation, pricing configuration, persistent storage setup, and audit submission, making it significantly easier for new infrastructure providers to supply compute resources to the Akash marketplace.

Developer Workflow Improvements: On Jan 22, Akash rolled out four new developer-specific workflows in Akash Console to improve user onboarding and application deployment. The new modes: Build & Deploy, Launch Container VM, Run Custom Container, and Upload SDL, streamline the deployment process based on technical skill levels and familiarity with Akash. The upgrade significantly reduces friction, particularly for developers accustomed to services like Vercel, DigitalOcean, or AWS.

AkashChat Enhancements: Following its initial rollout in early 2024, AkashChat was upgraded with support for DeepSeek-R1 and multiple open-source models (including LLaMA 3.1 and 3.3 variants) in Jan 2025. The AkashChat API now supports inference at scale for a range of advanced models, reinforcing the network’s ability to serve as a backend for permissionless conversational AI tools.

AI Pricing Transparency and Cost Savings: As of Q1 2025, Akash provided over 1,000 GPUs on its network, with more than 73% classified as high-density (e.g., H100, A100, H200). Pricing ranged from $0.76/hr for A100s to $1.93/hr for H200s.

Governance

Akash Network’s governance in Q1 2025 remained active with eight passed proposals addressing protocol upgrades, ecosystem analytics, community programs, and economic policy.

Community Programs and Ecosystem Support:

Proposal 277– Q1 Community Programs Funding (Jan 10): Allocated $92,500 in AKT to fund structured roles within the Akash community, including Vanguards (support), Navigators (technical onboarding), Alchemists (educational content), and Ringmasters (regional event organizers). These programs are intended to formalize contributor pathways and expand ecosystem participation through defined responsibilities.

Proposal 278– Q1 Community Activities and Events (Jan 10): Approved $158,575 in AKT to support hackathons, local meetups, and a community bounties program. The proposal emphasized supporting external event sponsorships and direct community execution of grassroots initiatives tied to cloud computing and AI.

Proposal 279– Q1 Events Resourcing: ETHDenver and NVIDIA GTC (Jan 22): Authorized $464,970 in AKT to fund Akash’s presence at ETHDenver (Feb 23–Mar 2) and NVIDIA GTC (Mar 17–22). The budget covered sponsorship, booth design, and community-hosted side events to increase visibility across both the AI and Web3 ecosystems.

Economic Policy and Treasury Adjustments

Proposal 280– Market Liquidity and Depth Enhancement (Feb 18): Allocated 3.5 million AKT from the community pool to engage external market makers and improve liquidity across major AKT trading pairs. The initiative aims to reduce volatility and broaden institutional access, shifting liquidity management responsibilities away from Overclock Labs and toward decentralized stewardship.

Proposal 282– Community Pool Tax Adjustment (Mar 13): Increased the community pool tax from 40% to 50%, redirecting more of the network’s inflationary rewards to fund ongoing ecosystem growth initiatives, such as provider incentives. The change responds to elevated onchain activity and resource usage metrics observed over recent quarters.

Proposal 283– Inflation Rate Reduction (Mar 14): Reduced the AKT token inflation range to reflect macroeconomic volatility and optimize issuance rates. The inflation maximum was lowered from 13% to 8%, and the minimum from 8% to 4%. The change is intended to support long-term token price stability and was passed with over 92% approval.

Network Upgrades and Data Infrastructure

Proposal 281– Mainnet v0.38.0 Upgrade (Mainnet 13) (Mar 11): Approved the upgrade of the Akash network to v0.38.0, introducing storage performance improvements via changes to x/store modules as defined in AEP-61. Validators were required to meet updated RAM requirements (96GB or equivalent), and the upgrade was coordinated using the standard Cosmovisor flow.

Proposal 284– Chakra Analytics, Indexing, and MCP Integration (Mar 19): Approved $16,200 in AKT to fund Chakra’s development of:

• A specialized Akash chain indexer for detailed metrics (e.g., GPU utilization, provider reliability).
• A public analytics dashboard.
• Integration with the Model Context Protocol (MCP) to allow AI agents to query Akash network data using natural language interfaces (e.g., Claude, Cursor).

All deliverables will be open-sourced and contributed to the core Akash repositories to support long-term transparency and agent-oriented infrastructure development.

Ecosystem and Recognition

Institutional Recognition: Grayscale included AKT in its Q1 2025 “Top 20 Assets to Watch,” citing its potential as a decentralized alternative to cloud incumbents. The recognition followed Akash’s inclusion in Galaxy Digital and VanEck reports covering the rise of sovereign AI infrastructure.

Akash Accelerate 2025: Akash hosted its second Akash Accelerate summit on June 23, 2025, in Brooklyn, following the inaugural flagship event held in Austin in May 2024. The annual gathering brings together contributors from across the decentralized AI and infrastructure ecosystem.

Closing Summary

In Q1 2025, Akash Network generated $1 million in lease revenue (+38% QoQ), despite a 24% decline in new leases to 46,000. The shift reflects a move toward longer-duration, higher-value workloads, supported by a 54% QoQ increase in average GPU usage and a 55% rise in average GPU capacity to 897 units. Average daily active leases increased 6% QoQ to 868, while the number of active providers grew modestly to 69 (+4% QoQ), aided by the launch of the Akash Provider Console.

Strategic developments included integrations with Envision Labs and NodeShift, both leveraging Akash for high-density GPU inference and training workloads. The network introduced autonomous AI agent deployments, a plugin for ElizaOS, and upgraded AkashChat, expanding its capabilities for decentralized AI infrastructure.

Governance activity remained active, with eight proposals passed covering inflation reduction, increased community pool tax, market liquidity support, and new indexing and analytics infrastructure via Chakra. These initiatives were paired with continued funding for community roles, events, and ecosystem growth.

Looking ahead, Akash plans to build on its momentum in GPU provisioning and decentralized AI infrastructure, following a set of key announcements shared during its Akash Accelerate summit in June, which underscored the network’s continued focus on enabling decentralized AI workloads.