Replit cuts Cloud compute, storage, and data-transfer prices up to 81%

Replit dropped prices across its Cloud platform on August 1, cutting Autoscale compute costs by more than 80% and database storage costs by more than 75%, with the new rates applied automatically to every existing deployment.

What's new

According to Replit's billing documentation, "Starting August 1, 2026, Replit Cloud prices drop across all plans" — Core, Pro, Teams, and Enterprise all receive the update with no action required: "The new prices apply automatically to all of your published projects at the start of your first billing cycle that begins on or after August 1, 2026."

The steepest cuts land on Autoscale deployments, Replit's serverless compute product:

• Autoscale compute units: $3.20 → $0.60 per million (a 81% cut)
• Autoscale requests: $1.20 → $0.40 per million (67% cut)
• Scheduled compute units: $3.20 → $0.60 per million, matching the Autoscale cut
• Outbound data transfer: $0.10 → $0.05 per GiB (50% cut)

App Storage also gets cheaper: data storage drops from $0.03 to $0.015 per GiB/month, with basic and advanced operations both cut by roughly a third. Database storage sees the largest single-line reduction, from $1.50 to $0.35 per GiB/month — a 77% cut — while database compute time holds flat at $0.16/hour: "Database compute prices are unchanged, while data storage prices decrease."

Reserved VM deployments get a more mixed update. Smaller configurations get cheaper (0.25 vCPU/1 GiB drops from $0.014 to $0.0104 per hour), but Replit is also consolidating its legacy and dedicated VM tiers into unified per-hour rates, which raises the price on at least one configuration (1 vCPU/4 GiB legacy goes from $0.042 to $0.0486/hour) as it's folded into the dedicated-tier pricing. Two new larger dedicated tiers — 8 vCPU/32 GiB and 16 vCPU/64 GiB — are introduced at $0.3611 and $0.7222 per hour respectively, where no equivalent option existed before.

Replit's documentation frames the change as fully automatic: existing deployments, databases, and storage "keep running without any action from you."

Context

Replit has iterated on its pricing model repeatedly over the past year, moving from flat compute pricing to effort-based pricing for its Agent product earlier in 2026, and restructuring its Core/Pro plan tiers before that. This latest change is narrower in scope — it doesn't touch Agent or subscription pricing, only the underlying Cloud infrastructure costs (Autoscale, App Storage, Reserved VMs, and databases) that sit under every deployed Replit app.

The cuts arrive as compute and storage costs across the industry have been falling — inference providers and cloud vendors have been passing down efficiency gains from newer hardware and denser infrastructure throughout 2026, and hosting platforms built on top of that infrastructure, like Replit, are following suit on their own margins.

Why it matters

For developers running always-on or usage-based apps on Replit, an 81% cut to Autoscale compute pricing meaningfully changes the cost calculus of hosting an AI agent or app that scales with traffic rather than sitting on a fixed VM. Combined with the 77% cut to database storage, teams running data-heavy Replit Agent-built apps get a substantial reduction in their standing infrastructure bill without having to migrate anything or change how their apps are built.

The move also underscores how much of Replit's competitive positioning now rests on the economics of running AI-built software cheaply at scale, not just on the coding-agent experience itself. As more of what gets built on Replit is AI-generated and deployed with minimal human tuning, the underlying hosting costs — not just the agent's per-task pricing — become a bigger part of whether building and running an app on the platform stays cheaper than doing it elsewhere.