Quasar

Introducing Subnet 24: Quasar by SILX AI We are solving the context window limitation by introducing a new subnet for competitive long-text understanding. What’s coming in 2026: - We are actively updating our subnet code - Quasar models will be released as open source - An RL framework for enhancing open-source models Quasar subnet leaderboard and ranking system We are backed by @dsvfund @bitstarterAI Founders : @TroyQuasar Eyad Gomaa @Farahatyoussef0 Youssef Farahat

Bittensor Ecosystem Highlights of the Week #59 // SUBNET UPDATES & ACHIEVEMENTS ➤ @vidaio_ SN85 Vidaio announced a joint venture with Pip Studios to bring VidaioOS into enterprise media workflows and major studio pipelines. (bit.ly/4ttZpHP) ➤ @QuasarModels SN24 Quasar announced their collaboration with @adaption_ai to generate high-quality adaptive datasets for training their foundation models and improving reasoning capabilities. (bit.ly/3RehewK) They recently opened their codebase and launched a global distillation competition. (bit.ly/49bP6k7) They also started a partnership with SN3 to support future Quasar model training once SN24 reaches its next performance target. (x.com/Farahatyoussef…) ➤ @actualinc SN95 Their new beta release is out, including support for AMD hardware on Windows and Linux. (x.com/Tom_A_Lynch/st…) ➤ @yanez__ai SN54 They added a third enterprise client to their Biometric Data product. (x.com/yanez__ai/stat…) ➤ @webuildscore SN44 Score came away from Refuel Forum in Barcelona with a multi country pipeline. (x.com/webuildscore/s…) They also added two new skills: ‘Detect Beverage’ and ‘Detect Crime’. (x.com/webuildscore/s…) ➤ @ridges_ai SN62 Ridges integrated Harbor by Terminal Bench to evaluate agents across more languages and more complex tasks. (x.com/ridges_ai/stat…) ➤ @leoma_ai SN99 Leoma just launched their AI video generation product. (x.com/leoma_ai/statu…) ➤ @resilabsai SN46 RESI has been accepted into the @Plaid Partner Program, bringing KYC and income verification directly into the platform. (x.com/resilabsai/sta…) ➤ @b1m_ai SN105 Beam introduced Prism, their scoring system for orchestrators. (x.com/b1m_ai/status/…) ➤ @chutes_ai SN64 AIxCrypto announced a strategic collaboration framework with Chutes AI. (x.com/chutes_ai/stat…) @jon_durbin is testing distributed pretraining of a 20B MoE model for under $10/hour across consumer GPUs. (x.com/jon_durbin/sta…) ➤ @trishoolai SN23 Trishool partnered with Velantris, an enterprise AI agents startup. (x.com/trishoolai/sta…) ➤ @babelbit SN59 Babelbit shared a live translation demo. (x.com/babelbit/statu…) ➤ @Bitcast_network SN93 Their new dashboard is live. (x.com/Bitcast_networ…) ➤ @LeadpoetAI SN71 Leadpoet has been selected for @Salesforce Launchpad. (x.com/LeadpoetAI/sta…) ➤ @SwarmSubnet SN124 Swarm has been selected for Enlaira, Andorra’s official startup acceleration program. (x.com/SwarmSubnet/st…) ➤ @theminos_ai SN107 Minos is now live on mainnet. (x.com/theminos_ai/st…) ➤ @vocence_bt SN78 Vocence Studio is now live. (x.com/vocence_bt/sta…) ➤ @0x_Markets SN35 They unveiled their new website. (x.com/0x_Markets/sta…) // SUBNET INVESTMENTS ➤ @VantaTrading SN8 @stillcorecap has invested in Vanta’s SN8 alpha token. (x.com/VantaTrading/s…) // NEW SUBNETS ➤ @heydittoai SN118 Ditto aims to be the open source version of Claude Cowork on Bittensor. (x.com/heydittoai/sta…) // BITTENSOR ECOSYSTEM ➤ @taostats They introduced Bittensor Auth Gateway. (x.com/taostats/statu…) ➤ @bitstarterAI They partnered with @TargonCompute to bring confidential compute to teams in their Machine Learning track. (x.com/bitstarterAI/s…) ➤ @Rubicon_Bridge Rubicon Bridge V2 is live. (x.com/Rubicon_Bridge…) ➤ @YumaGroup Institutions can now stake $TAO directly to Yuma’s validator via Colossus Hub. (x.com/YumaGroup/stat…) // PODCASTS & ARTICLES ➤ @11AMdotclub podcast with @Tom_A_Lynch from Actual. (x.com/11AMdotclub/st…) ➤ @EyeOn_AI podcast with @macrocrux from @MacrocosmosAI (x.com/MacrocosmosAI/…) ➤ Hash Rate by @markjeffrey with @0xarrash from Vanta. (x.com/markjeffrey/st…) ➤ @JesusMartinez podcast with Mark Jeffrey (x.com/JesusMartinez/…) ➤ Jesus Martinez podcast with @Sebyverse from Resi (x.com/JesusMartinez/…) ➤ Revenue Search 63 by @SiamKidd and @MarkCreaser with @NiomeAI (x.com/dsvfund/status…) ➤ @VenturaLabs podcast with @SomaSubnet (x.com/VenturaLabs/st…) ➤ @GSchvey “Bittensor state of the art” thread. (x.com/GSchvey/status…)

SN 3 X SN 24 Building real intelligence accross the network

@QuasarModels @adaption_ai this is like when we first hear Anthropic partnership with ScaleAI before they deliver Claude Opus Something huge is coming through!

When we said we are going to deliver state of the art, we meant it.

I’m happy to announce our collaboration with Adaption Ai

The history of decentralized training runs has mainly focused on large-scale execution, as this is important for open and transparent training. But they’ve missed one critical thing: quality. Even the most successful training runs, like SN3, are technically unusable in practice because quality was never the goal. At Quasar SN24, we’re changing that. We are building the largest MoE training run with SOTA performance and quality as core priorities. This means the outputs won’t just be impressive decentralized runs they will be SOTA long-context models that are actually usable.

Happy to announce our collaboration with @adaption_ai Adaption Labs is an AI research company focused on building adaptive intelligence systems . Through this partnership, Adaption Labs will provide SILX AI with state-of-the-art adaptive data to support the training of the Quasar foundation models. Their role will be to generate and refine high-quality, adaptive datasets at scale, enabling Quasar to continuously improve its reasoning and generalization capabilities. This collaboration strengthens Quasar path toward achieving SOTA performance and competing with leading closed-source models. The company is co-founded by Sara Hooker, former Vice President of Research at Cohere and a veteran researcher from Google DeepMind, alongside Sudip Roy. Adaption Labs has also raised $50M in seed funding to advance its mission in adaptive AI.

YOOOO #SN24 $TAO ! @QuasarModels

🚀 Quasar Competition Starts Now Miners, this is your moment. The Quasar codebase is now fully open. We are launching a global competition to distill knowledge into Quasar and push it to state-of-the-art performance. [🏆 Winner takes it all ] The first training run begins now Anyone can participate bring your best methods, your best ideas. For details, updates, and how to get started, join the Quasar Discord channel on Bittensor. Let’s build 👇

@GuardianX59358 @QuasarModels not sure how to say it but its not even our main event

First announcement of April and not the last. This release is important. SN24 will train the Quasar architecture that solves long context how do we know? Coming Quasar tops OpenAI on MRCR, harder 8 needles at 1M context

@JackDersoy not a new team member a new org *

@QuasarModels And who’s the new team member? 🤔🤯 @QuasarModels

First announcement of April. We introduce Quasar-3B (1B Active) a looped continuous-time transformer built for long-context intelligence Quasar uses a hybrid architecture combining Quasar layers + GLA, enabling efficient stateful reasoning with stable long-range dependency handling. This release is the base model of the Quasar system, designed for distributed training and distillation at scale. In the coming days, SN24 miners will begin working with us to distill knowledge from frontier models like Qwen into Quasar, pushing toward a new SOTA in long-context modeling. The training roadmap is staged: Stage 1: Quasar-RoPE stable pretraining with 16K context, establishing the base representation for the system Stage 2: Quasar-DroPE continued training with distillation as the core mechanism, removing positional encodings and scaling toward 5M token context In both stages, distillation remains the central training signal, driven collaboratively through SN24 miners using knowledge transfer from frontier models. Read more about Quasar and get ready for mining 👇

@JackDersoy Not the last update for this month

@QuasarModels This is great news! Keep the updates flowing! @QuasarModels

@CryptoZPunisher An organization, not a person.

Bittensor >> $dTAO Subnet 24: Quasar @QuasarModels silxinc.com I’ve read some interesting things on your Discord, so tell me more. Why are we not ready for what’s coming? And who is this person joining your team?

Cooking hard 🧑‍🍳