表现
RD 熱度指數
RD 增長指數
| 1h | 24h | 7d | 30d | 60d | 90d |
|---|---|---|---|---|---|
| 0.25% | 0.68% | 9.24% | 23.40% | 4.04% | 19.31% |
比較
代幣經濟學
解鎖進度
數據源: CoinMarketCap
TAO is the native token of the Bittensor decentralized AI network, using a bitcoin-like model with all issuance via miner/validator rewards. Capped at 21 million, it halves stepwise by cumulative issuance, used for staking, fees and dTAO subnet liquidity.
開放市場
| # | 交易所 | 交易對 | 價格 | +2% 深度 | -2% 深度 | 成交量(24小時) | 成交量% |
|---|---|---|---|---|---|---|---|
| 1 | KKuCoinCEX | TAO/USDT | $231.9 | $380,867 | $473,734 | $23,960,687 | 20.07% |
| 2 | BBiFinanceDEX | TAO/USDT | $231.8 | $768,778 | $704,275 | $14,295,118 | 11.97% |
| 3 | BBinanceCEX | TAO/USDT | $231.76 | $285,521 | $626,575 | $12,006,323 | 10.05% |
| 4 | BBTCCDEX | TAO/USDT | $232 | $1,949,661 | $2,634,059 | $5,965,358 | 5.00% |
| 5 | TAO/USD | $232.08 | $225,306 | $619,774 | $5,928,425 | 4.96% | |
| 6 | WWhiteBITDEX | TAO/USDT | $232.28 | $799 | $68,749 | $5,655,199 | 4.74% |
| 7 | CCoinWDEX | TAO/USDT | $232.28 | $110,929 | $328,611 | $4,517,693 | 3.78% |
| 8 | KKrakenCEX | TAO/USD | $231.94 | $384,884 | $605,730 | $4,404,114 | 3.69% |
| 9 | OOKXCEX | TAO/USDT | $232.08 | $43,569 | $180,592 | $3,822,735 | 3.20% |
| 10 | WWEEXDEX | TAO/USDT | $231.77 | $554,284 | $582,665 | $3,798,358 | 3.18% |
跟隨更新
跟隨列表
人物
關於Bittensor
Bittensor is a decentralized machine-learning network running on its own blockchain (the Finney mainnet, built on the Substrate framework), serving researchers and developers who want to train, provide and consume AI models in an open way. It splits the network into subnets, each an independent incentive market: miners produce intelligence such as inference, data or predictions, validators score contributions under their own rules, and Yuma consensus aggregates them at the root layer. Its core problem is how to keep incentivizing open-source intelligence without a centralized company; concretely, Dynamic TAO gives each subnet an alpha token and an AMM pool while Taoflow ties TAO emissions to net staking inflows, so issuance follows real capital rather than root-validator votes, and subnet participants must post miner collateral to meet the Emission Gate threshold.
This market-driven emission design lets subnet quality be set by capital rather than politics, at the cost of complex protocol logic and a steep learning curve for ordinary users, plus TAO's heavy stake concentration making security rely on a few large stakers. In April 2026, Grayscale and Bitwise filed spot TAO ETF applications with the SEC, bringing the network into the regulatory purview of traditional finance for the first time.
Bittensor 具有長期意義的關鍵里程碑包括：灰度於 2024 年首次為 Bittensor 信託（GTAO）提交 ETF 申請，標誌著 TAO 首個機構合規通道的建立。Upbit 於 2025 年上線 TAO，擴大了亞洲主要零售市場的准入。OKX 推出的 TAO/USDT 現貨交易進一步深化了全球流動性。DCG 子公司 Yuma 啟動專項生態基金，顯示機構資本對子網經濟的長期承諾。2025 年 4 月 Covenant AI 子網退出引發治理危機，暴露出所有者缺乏利益綁定的問題，促使共同創辦人提出鎖倉質押機制（BIT-0011）以協調長期激勵。隨後發布的 18 個月去中心化路線圖，概述了從核心團隊控制向社群治理的過渡，包括 dTAO 機制與排放減半計畫。這些事件共同代表了 Bittensor 從實驗性獎勵網路向市場驅動的 AI 生產基礎設施的演進，其中機構參與、交易所整合與治理改革構成了其長期發展的結構性支柱。
Institutional and regulatory: On April 2, 2026, Grayscale and Bitwise both filed spot TAO ETF applications with the SEC (Grayscale seeking to convert its Bittensor Trust into a NYSE Arca-listed ETF); the decision window was expected in August, but no approval had been granted as of October. On April 7, Grayscale raised TAO's weight in its AI fund to 43.06%, and on October 15, a Grayscale–Republic decentralized AI fund launched with TAO at 29.15%.
Protocol upgrades: The v437 upgrade (July 20–26) introduced miner collateral requirements; v440 (July 28) activated the Emission Gate, requiring subnets to meet a minimum demand threshold for full emissions; and Null Consensus (October 16) expanded subnet capacity from 256 to 2,500 participants and added optional proof-of-work registration, lifting TAO more than 10% above $285 that day. The second halving is scheduled for December 2026, cutting daily emission from about 3,600 TAO by half again.
Fundamentals: Q1 network AI-related revenue was about $43 million, still below annualized emissions worth over $148 million; about 70% of supply is staked at roughly 8%–11% APY. In late September, TAO briefly traded near $320 (up about 50% on the week), with market cap back near $3.6 billion.
In May 2026 the Robin τ upgrade raised the subnet cap from 128 to 256 and bridged canonical TAO to Solana through Wormhole; Taoflow replaced price-based emissions with real-time net staking inflows, zeroing emissions for subnets with sustained outflows. Covenant AI exited the same period over a governance dispute that touched its 72B-parameter model and subnet incentive splits, and TAO fell more than 20% in a day. First-quarter revenue from real AI compute was about $43 million.
On October 7, Runtime 475 added an optional Null Consensus mode in which a single largest-stake validator sets weights, lifting participants per subnet from 256 to 2,500 and adding an optional proof-of-work registration path; on October 20, eTAO let subnet owners set custom splits of alpha-token emissions among owners, validators and miners. A September 15 issuance of 172,001 TAO to close an accounting gap moved the next halving about 48 days closer. Grayscale and Bitwise jointly filed for a spot ETF in April, TAO made up 29.15% of the Grayscale-Republic Decentralized AI Fund on October 15, and about 64.5% of TAO is staked.
Bittensor's Future strategy and roadmap: Bittensor's mechanism experiments matter more than its roadmap: within a year emissions went from price signals to net staking inflows and then, in late June, back to an EMA-price basis layered with miner-burn penalties and active-mining gates, showing the team is still tuning parameters to balance subsidy against real compute. Network-wide AI compute revenue reached $43 million in Q1 2026 and 128 active subnets are valued near $1.5 billion combined, yet independent estimates put Chutes at 1.6 to 3.5 times the cost of centralised alternatives once emission subsidies are stripped out, so demand is still paid for by inflation. With the halving cutting daily emissions to 1,800, whether a subnet covers its own costs becomes the only survival test, not how many seats it registers.
雅各布·羅伯特·斯蒂夫斯（創辦人）：曾任Google機器學習工程師，於2015年構思Bittensor，並於2021年1月透過Opentensor基金會啟動主網。他於2025年辭去執行長職務以加速去中心化，但仍積極參與開發。 阿拉·沙巴納（共同創辦人）：擁有應用人工智慧博士學位，曾於VMware（2018-2019年）從事分散式運算工作，並共同創立Bittensor與Opentensor基金會，自2019年起主導協議開發。他亦共同創立了Crucible Labs。