Most Bittensor Subnets Will Die – These 4 Will Survive

A new breakdown by TAO Templar cuts through one of the most burning concerns: are owners burning their own emissions?

The Core Argument

Most subnets won’t be around in six months. The reason is mechanical: liquidation prices across all subnets drop roughly 2% per week, meaning the risk of holding any alpha token compounds steadily. By mid-2026, the average liquidation price across the network will be roughly halved, putting real pressure on subnets that aren’t actively driving value into their token.

Tao Templar used a unique filter. Forget partnerships, forget roadmap promises, forget influencer shills. The filter that actually matters is whether subnet owners are taking real money and burning it.

Why burns are the cleanest signal:

• They can’t be forged
• They cost real money
• They permanently reduce supply
• They prove conviction without requiring trust in the owner

The Four That Make the Cut

1. Chutes: The Revenue Flywheel

• Multi-model inference platform (LLMs, image, video, text-to-speech)
• Consistently the cheapest inference provider on Bittensor
• Manual burns are running steadily and growing
• Revenue now offsets over 50% of miner emissions, up from ~40% a few months ago
• Trajectory: full miner-emission offset

2. Lium: The Most Conscientious Subnet on the Network

A GPU marketplace, the team meticulously plugs every leak in their token economy.

• 95% of owner emissions burned (only 5% kept for development)
• High miner burn rate alongside it
• Commit-reveal enabled to combat weight copying
• Multiple subnet-owner UIDs used as additional burn vectors against weight copiers

The one knock is that the subnet owner, Fish, isn’t doxxed. But Tao Templar counters: don’t trust, verify. The burns are on-chain and undeniable, which removes most of the trust assumption.

3. Bitcast: First to Fully Offset Miner Emissions With Revenue

• Matches creators with brands; miners are content creators, so hard costs are minimal
• First subnet to fully offset miner emissions with revenue
• Commit-reveal enabled
• High miner burn rate keeps emissions tight
• Every dip in the burn chart represents an actual brand campaign being funded

The weakness: burns happen in lumpy batches rather than the consistent daily cadence that Chutes and Liam have established. The owner has signaled plans to lock the majority of the owner-key stake and revenue once Bittensor’s locking mechanism ships.

4. Gradients: Smaller Burns, Right Direction

An auto-ML platform popular with Bittensor OGs.

• Two manual burns in the past year, ~6,000 alpha each (~$100K in token value)
• Not yet revenue-driven; team has stated buybacks scale up alongside B2B sales
• Sales motion still early-stage

The burns are modest, but they’re happening, and they imply an owner thinking about token value rather than extracting from it.

Why This Framework Matters

The deeper point is that most subnet owners optimize the wrong side of the emissions pie. Emissions split roughly 41% to miners, 18% to subnet owners, 41% to validators and stakers. Most teams obsess over miner burn and ignore the other two-thirds entirely.

Lium is the exception. Burning owner emissions, defending validator emissions from weight copiers, and pushing toward covering miner emissions with revenue. That’s what a complete flywheel looks like.

The Locking Wildcard

Tao Templar flags Bittensor’s upcoming conviction locking mechanism as the next major sorting tool. Where burns are the gold standard, irreversible, no trust required, locks are a softer but still meaningful indication. Owners commit their alpha for a fixed period before being able to sell. Once locking ships, the question is which teams are utilizing the conviction lock mechanism.

The Takeaway

Two things to watch as the network matures:

• Burns are the only indication that doesn’t lie. Anything else, like partnerships, deals, future revenue, etc., requires trust. Burns require money.
• Risk on alpha tokens is rising 2% per week. Subnets that aren’t actively offsetting that decay with revenue, burns, or both will face deregistration pressure.

Chutes, Lium, Bitcast, and Gradients are the four that Tao Templar is willing to bet survive it.

Watch the full video below: