Braiins has released Manager Agent 4.13.0, a fleet-management update aimed less at flashy tuning features and more at something industrial Bitcoin mines depend on every day: keeping device identity, telemetry, and control accurate even when miners are reflashed, APIs return bad data, or firmware changes underneath the management system.

The September 17, 2026 release adds automatic firmware-type correction, automatic miner re-detection when a device stops returning valid data, faster data delivery for small and midsize farms, and improved handling of malformed Antminer API responses that could previously leave gaps in telemetry.

The Agent Sits Between the Fleet and the Cloud

Braiins describes the Manager Agent as the local communication layer between mining hardware and the Braiins Manager backend. It runs on a server inside the site network, polls miners for operating data, and passes commands back to the fleet.

That means the Agent is not the same thing as Braiins OS firmware. Firmware runs on the miner itself. The Manager Agent runs on a local Windows or Linux host and handles monitoring, bulk actions, pool configuration, pausing, resuming, and other fleet-level operations. Braiins recommends Ubuntu or Debian for the most stable Agent deployment, although Windows 11 is also supported.

For a site operator, that architecture makes the Agent a bridge between individual ASIC APIs and the remote management dashboard. If the bridge misidentifies a miner, holds stale firmware information, or stops parsing telemetry correctly, the problem can spread upward into dashboards, automation rules, maintenance tickets, and curtailment logic.

Firmware Type Can Now Correct Itself During Polling

The most important change in 4.13.0 is automatic firmware-type correction. Braiins says firmware changes previously could require manual intervention in the user interface or remain incorrect until a later scan. The new Agent attempts to detect those changes during ordinary data polling and correct the device type sooner.

That matters on mixed fleets where miners may move between stock firmware and third-party images during testing, troubleshooting, or performance optimization. BitcoinVersus recently covered why firmware changes on newer Bitmain hardware can affect third-party installation workflows. A management platform that recognizes those changes quickly reduces the chance of operators making decisions against stale device metadata.

Bad Miner Data Can Trigger Automatic Re-Detection

Braiins also changed what happens when a miner stops returning valid data. The Agent can now automatically re-detect the device, including cases where a miner was reflashed outside Braiins Manager.

In practical terms, that is a self-healing behavior. Instead of assuming the old device identity and API behavior are still valid forever, the Agent can re-evaluate the miner when its responses stop making sense. That is especially useful at sites where field technicians may flash or replace control boards independently of the management server.

The same release improves handling of malformed Antminer API responses. Braiins says bad responses could previously cause missing telemetry on some devices. For large fleets, missing data is not just a dashboard nuisance. It can hide temperature, power, hashrate, pool, or device-state problems long enough to delay maintenance.

Faster Batching Targets Small and Midsize Farms

Manager Agent 4.13.0 also improves batching so data arrives faster for small and midsize operations. That follows earlier Agent updates focused on command-batch resilience, including version 4.12.0, which improved processing of large command batches during events such as fleet ramp-up after curtailment.

That sequence shows how management software is becoming part of the operating system of the mine. A site may have excellent network and PDU monitoring, but the ASIC management layer still has to keep thousands of miner-specific data points synchronized quickly enough for operators and automation to trust them.

📣 Braiins Manager is here 📣Our free mining management system is open to the public. Sign up today for free to maximize your mining operation’s efficiency.
byu/Braiins_mining inBraiins

The Release Also Adds Z15-Family Support

Version 4.13.0 adds support for Antminer Z15, Z15j 320 kSol/s, and Z15 Pro 860 kSol/s models. Those are Equihash miners rather than Bitcoin SHA-256 ASICs, so this is not a new Bitcoin-miner hardware compatibility announcement. It does, however, show that Braiins Manager is being developed as a mixed-fleet operations platform rather than a Bitcoin-only device list.

That matters for hosting providers that manage multiple algorithms or customer-owned hardware under the same operational team. A single management layer can reduce the number of monitoring and ticketing systems technicians need to learn, provided the platform correctly identifies each device and exposes the right controls.

Fleet Automation Is Only as Good as Its Telemetry

Braiins Manager can automate actions such as rebooting underperforming miners, changing pool settings, pausing machines, and responding to power-market conditions. BitcoinVersus previously covered Braiins Price Adapt, which automatically changes ASIC power targets as mining economics and energy prices move.

But automation depends on trustworthy state. If a miner is identified with the wrong firmware, reports broken API data, or disappears from telemetry after a manual reflash, a sophisticated automation rule can still make the wrong decision. Manager Agent 4.13.0 is therefore less about adding a new control and more about making the existing control stack harder to confuse.

Watch Braiins Manager Control a Mining Fleet

Braiins’ Manager overview below shows how the platform ties fleet monitoring, automation, curtailment, and remote control together through the local Agent.

What Operators Should Check After Updating

• Confirm miners that were recently reflashed are identified with the correct firmware type.
• Watch for miners that previously showed missing or intermittent telemetry.
• Verify pool, hashrate, temperature, and power data after the Agent update.
• Test bulk commands on a small device group before applying them fleet-wide.
• Confirm automation and curtailment rules still target the intended workers.
• Keep the local Agent host healthy, patched, and reachable from the miner network.

For large Bitcoin mines, the management plane has become almost as operationally important as the individual miner firmware. Manager Agent 4.13.0 is a small release by feature count, but its focus on re-detection, data correctness, and faster polling targets the kinds of quiet failures that can become expensive when multiplied across thousands of ASICs.