Now live: SN64 @chutes_ai is the leading serverless AI compute network on @bittensor, powering trillions of tokens per month across open-source LLMs and every open-source modality, from image and video to speech and music. Start trading today → app.kraken.com/JDNW/SN64

Not all partnerships are created equal.. some open doors... others open worlds! We are proud to announce: @manakoai customers will soon be able to cut bandwidth costs with a single click. @vidaio_ 's compression models, natively integrated. Every camera intelligent AND optimised As Manako grows, Vidaio grows with it This is optimisation on the edge Real production, real workflows, real customers.

We just let the first 50 on the waitlist in. gm is live on mainnet and the first waitlist members have access right now. Check your inbox. Every account got $10 of credit on us. Go break something.

Quick update: Forge testnet has been a much stronger signal than expected. What started as a testnet release turned into hundreds of users, heavy traffic, and direct community interaction with the core primitive: borrowing TAO against Alpha collateral. That traction shows real curiosity and demand around Bittensor-native credit, and it gives the team live feedback on how users interact with Alpha-backed lending before broader rollout. @ForgeLending is the first proving ground for this market. It is where demand becomes visible, user behavior gets tested, and the risk questions become concrete: which Alpha markets should be accepted as collateral, how much exposure is safe, and how quickly should parameters adapt as liquidity, volatility, and drawdowns change? That is where Endure comes in. Endure is being built as the risk-intelligence layer behind Forge, but the long-term vision is broader than one lending market. The goal is to create an intelligence market where risk models compete, get benchmarked, and improve over time. In that world, raw market data is not enough. The valuable output is evidence-backed judgment: which signals matter, which models were early, which recommendations were accurate, and which insights were actually useful when conditions changed. For Forge, the current work is moving toward a lending V1 path where miners will be able to submit structured Alpha collateral risk recommendations and validators will be able to score them against realized market behavior. Recent progress has focused on turning that path into an executable subnet loop: Forge lending registry selection, reveal validation, Alpha collateral whitelist gates, lending market-data fixtures, real mainnet drawdown windows, miner submission flows, lending consensus logic, and validator hardening around state, pacing, config, and schema handling. Over the next phase, Endure should move closer to producing the types of outputs a lending market actually needs: safer pricing inputs, collateral-factor guidance, risk tiers, cap recommendations, liquidation assumptions, and eventually more adaptive risk parameters for Forge. At the same time, the team is working on the harder research problem: how to benchmark risk intelligence itself. That means thinking in terms of SOTA, baselines, and measurable improvement. Not just “did the model produce an answer?” but “did the model beat a naive benchmark, did it stay calibrated under stress, and did it turn noisy information into a decision that held up against future outcomes?” This is where concepts like Validated Insight Rate become important: measuring how reliably raw information turns into evidence-backed decisions. The potential is a network that can identify which models are consistently early, accurate, calibrated, and economically useful. Forge showed demand far beyond expectations for Alpha-backed TAO borrowing. Endure is building the market that can score the risk behind it.

This Thursday on Novelty Search :: Bittensor SN54 Yanez @yanez__ai shows how #SN54 uses Bittensor miners to generate synthetic adversarial identity data for financial crime prevention, helping compliance teams stress-test sanctions screening, fraud detection, and KYC systems before criminals do. Join live via Bittensor Discord. Hosted by @const_reborn

ANNOUNCING: Djinn has decided to open source all its code! Including code for miners and validators on Bittensor SN103, the website, app, smart contracts, tests, and more! As of five months ago. Yes, day one. We were the people’s subnet back then too, we just didn’t know it yet.

A handful of companies control most of the world's blockchain RPC traffic and set their own margins. We built a marketplace where independent operators compete that price down instead. Inaugural blog post is live 👇 blockmachine.io/blog/introdu…

We’re excited to announce that Trishool’s HaloGuard 1.0 𝐡𝐚𝐬 𝐚𝐜𝐡𝐢𝐞𝐯𝐞𝐝 𝐒𝐎𝐓𝐀 prompt-safety performance among open-weight guard models. Today, we present HaloGuard 1.0, a constitutional input classifier for multilingual AI safety. It is built as a first-layer input guard that checks user prompts before they reach a downstream LLM, agent, or application. This is part of the safety infrastructure being built through @trishoolai , our decentralised AI red-teaming subnet on Bittensor SN23. Full arXiv paper goes live soon.

Our new leaderboard is out! leaderboard.perturbai.io/ Now miners can view performance across the network, compare results locally, and independently audit the subnet. #sn26 #perturb #bittensor bittensor:native

The @opentensor Foundation has long run the free RPC service that the Bittensor ecosystem depends on. It's time for SN19 to take that on. We're proposing a redesigned free tier - no account required, no API key, rate-limited by IP. Powered by miners who lock up alpha as conviction. The mechanism is elegant: to serve free traffic, miners stake skin in the game. Spamming the free tier for incentive and dumping means dumping on yourself. Bittensor gets a free & reliable RPC, run by its own community. We're putting this to an owner vote. Vote here 👉 governance.taostats.io/p/sn1…

AI agents can already work, create, and earn. Now they have a marketplace. Introducing OKX AI: where agents discover work, hire each other, complete tasks, and get paid onchain. The one-person company just got an agentic workforce. Start here: okx.ai

AI can work with sensitive customer data without actually seeing it. In this chat, GM’s privacy mechanism lets the assistant answer correctly and reference private info, while the underlying email addresses remain hidden from the model. Privacy + utility 👇

Here is the full version of my Agent workshop covering my history with agents and current setup, from @proofoftalk. Accompanying materials an info here: s3.hippius.com/poppy/share/p… *full disclosure, I am getting a bit annoyed with Openclaw recently and going to do some tests with Hermes, will let you know how it goes.

gm, day one's done and gm held up. four miners running now, serving every frontier model, plus open-weight end to end in the enclave via chutes. In production, not a diagram. here's a free api key

ForeverMoney is no longer 98 We decided to separate the ForeverMoney Automated Liquidity Management infrastructure we built - and will continue building and operating - from Subnet 98. What this means: ▸ For ForeverMoney users: All services of ForeverMoney, Automated Liquidity Management on Aerodrome, as well as price alignment services between Finney and Base, will keep operating as they do today. We have been the top miner subnet; service quality will remain unaffected. ▸ For Alpha Holders: Before initiating the cold-key swap, 100% of owner-alpha was locked in Conviction in perpetuity. The new subnet owner will take over in roughly 5 days. Why we made this decision: We believe Bittensor can give you two types of superpowers: 1. Scaling supply side - Solving the chicken-and-egg problem. Who came first - GPUs or clients? Without clients you cannot rent GPUs; without GPUs you have nothing to sell. Alpha emissions solve this by bringing supply online before scaling demand (Chutes example). 2. Scaling intelligence - Tapping into decentralised R&D fueled by market competition. Building state-of-the-art models for computer vision or 3D generation is a heavy lift for a small startup. Alpha emissions solve this by enabling us to leverage a globally competing workforce (Score / 404gen example). However, we feel it's not an ideal fit for on-chain liquidity management: 1. Scaling supply side - We simulated renting/buying liquidity with alpha tokens via yield farming or bonding programs, but decided against it. Renting liquidity via bribes rarely leads to sticky liquidity. Building protocol-owned liquidity by selling alpha would have created substantial sell pressure. 2. Scaling intelligence - After intensive research and simulations, we came to accept that Impermanent Loss (the cost of providing liquidity on DEXs) is more of a feature than a bug. Marginal improvements are possible, but there are hard limitations. Adding decentralized intelligence won't turn a race horse into a race car. Beyond this, there are other complications when applying decentralized intelligence to financial strategies - attack vectors, speed, and scalability concerns. What's next: We're staying in the ecosystem and contributing to the important mission of growing on-chain accessibility of TAO beyond Finney. Stay tuned for updates. Thanks for your trust and support. - ForeverMoney Team

🔴 BITTENSOR TAO DEBATE (First 19 minutes is a Bittensor overview, feel free to skip ahead if that's not needed) Thank you @mogmachine for your time $TAO $ICP $CLOUD ☁️♾