Compelle is Bittensor subnet 82 (SN82), a project in the Other (Adversarial Reasoning and Debate) lane. On-chain debate, ranked by Elo
Compelle turns Bittensor subnet 82 into a public arena where AI models argue opposite sides of real questions, and the strategy, transcript, and Elo result of every debate are written on chain.
It is an educational deep dive, not a recommendation to stake. For live price, pool depth, and performance, always open the live subnet page.
What is Compelle (SN82)?
Compelle is a Bittensor subnet (number 82) that runs structured debates between AI language models. Someone picks a question, two AI models are assigned to argue the Pro and Con sides, and the system scores which side actually made the stronger case. Winners and losers move up or down a public Elo ranking, the same idea used to rank chess players.
In plain English: It is a competitive debate league for AI, with an open leaderboard. Models climb by winning arguments, not by claiming they are smart.
Centralized reference points: No direct equivalent; closest is private AI-lab debate evaluations
How miners and validators work
Like every Bittensor subnet, SN82 is a two-sided market: miners produce work, validators score it, and emissions follow those weights under the network’s dTAO economics.
- Miners commit a written debate strategy on chain: the playbook an AI debater follows when it argues a side.
- Validators pair those strategies in Pro and Con debates on a shared topic each epoch, run an AI judge to decide the winner, update Elo, and set weights from the results.
Mechanism and design
The mechanism is described in the subnet's own validator repository, which the current slot owner declares on chain. Miners do not stream live work. Instead each miner commits a written debate strategy on chain, a set of instructions an AI debater follows. Once per epoch, every validator reads the metagraph and those commitments, then runs a round-robin tournament: it pairs eligible miners in Pro and Con debates over a shared topic, has an AI judge decide each matchup, updates an Elo score, and calls set_weights from the result. Every validator runs the same configuration, which the repository treats as a shared consensus contract, and writes a signed record of each epoch to a public endpoint.
The debate topics are not random trivia. They are drawn from prediction markets such as Polymarket and Kalshi, framed as real propositions ("Bitcoin will hit a given price by a given date"), so the debaters argue questions with stakes and a public resolution. The language-model inference for both the debaters and the judge is offloaded to Chutes, itself a Bittensor compute subnet, so Compelle leans on one part of the network to power another.
A few snapshot readings to place the subnet, as of mid-June 2026. The alpha price sits near 0.00663 TAO, the subnet draws roughly 0.8 percent of network emissions, and the liquidity pool holds about 1,033 TAO of root depth. Compelle is still inside its four-month network immunity window, which runs until roughly late August 2026, so it cannot be deregistered yet regardless of where it ranks. A high root proportion in the pool is consistent with a subnet this young, whose price has not settled.
Why the project exists
- The problem it solves: There is no open, continuously running benchmark for adversarial reasoning, the skill of holding up an argument when something is actively trying to tear it down. Labs test this privately, but the results stay behind closed doors.
- The opportunity: A public, always-on record of which reasoning strategies actually win under pressure. Compelle's own framing is that argument under fire may be a path toward more general intelligence, and it makes that contest measurable.
- The Bittensor advantage: Because the strategies, transcripts, and every Elo shift are open and on chain, the ranking can be checked rather than taken on trust. Miners are paid for finding genuinely stronger strategies, not for a leaderboard nobody can audit.
- Traction signals: The product is live at compelle.com, where the site reports tens of thousands of judged debates run continuously on SN82. The validator repository is active, with 67 commits and a most recent push on June 5, 2026. Independent social coverage is still thin, so the live site and the repository are the main evidence of activity.
What stakers should check before buying alpha
Subnet storytelling is cheap. On-chain and product checks are not. Use the same framework as our subnet metrics guide:
| Check | Why |
|---|---|
| Real product / docs / repo | Separates shipping teams from narrative-only slots |
| Pool depth (TAO) | Your exit door: thin pools amplify slippage |
| Net flows (7d / 30d) | Under flow-based emissions, capital votes with its feet |
| Emission share | Whether the market is funding this subnet right now |
| Ownership concentration | Large holders can move a small alpha market |
| Immunity / deregistration | Post-immunity weak slots can lose the race (see immunity guide) |
Pull the live numbers on Compelle · SN82 rather than trusting any static screenshot in a blog post.
Risk factors
- New subnet, single public maintainer: The validator repository shows one contributor and was created in May 2026. The design is recent and unproven at scale, and the public codebase currently rests on a single maintainer.
- Market and price volatility: The alpha token's price has fallen sharply since launch (snapshot as of June 12, 2026). As a young subnet whose price has not settled, it can move hard in either direction.
- Emission and staking flows: Net staking flows were negative over the trailing week at the snapshot. Under the Taoflow model, sustained net outflows lower a subnet's share of emissions over time. Deregistration is off the table during the immunity period, but becomes a live consideration afterward if emission share falls toward the bottom of the network.
- Liquidity: Pool depth is modest, around 1,033 TAO at the snapshot, so larger stakes or unstakes will face slippage.
- Inference dependency: Debates and judging run on Chutes. The arena depends on that external inference provider staying available and affordable.
How this fits a TAO portfolio
Many holders keep a root vs subnet split: root for calmer TAO exposure, subnets like Compelle for thematic upside (and higher variance). If you use TaoScale, you can size SN82 from your wallet without handing over custody. See non-custodial staking.
- Live market: SN82 on TaoScale
- Browse and stake: /invest
- More subnet explainers: /blog/subnets
FAQ
What does Compelle do on Bittensor?
It is a competitive debate league for AI, with an open leaderboard. Models climb by winning arguments, not by claiming they are smart.
What is the SN82 alpha token?
When you stake TAO into subnet 82, you receive that subnet’s alpha. Its price versus TAO moves with the subnet’s pool (see what is dTAO).
How do I research Compelle before staking?
Read the mechanism above, open the official repo/docs when available, then verify pool depth, flows, and activity on SN82. Cross-check with the metrics checklist.
Is this financial advice?
No. TaoScale publishes educational explainers and live tools. Decisions to stake are yours alone.
Not financial advice. Educational overview of Compelle (SN82) compiled from public project materials and ecosystem documentation. Mechanisms and market structure change, so verify live data and official sources before acting.
