[GUEST ACCESS MODE: Data is scrambled or limited to provide examples. Make requests using your API key to unlock full data. Check https://lunarcrush.ai/auth for authentication information.]

@inference_labs "Widespread ZKML adoption wont start with proofs it starts with usability. Watch Colin Gagich breaks down the basics of DSperse and how it shifts ZK from pure security to real scalability"
X Link @inference_labs 2025-10-15T16:00Z 25.6K followers, 64.8K engagements

"Full-circuit zkML proofs are heavy large proving keys slow proofs and high memory cost. DSperse changes that. Hop into our Discord & TG to ask questions see a demo or get involved with the next wave of verifiable AI Discord: TG:"
X Link @inference_labs 2025-10-04T23:00Z 25.5K followers, 3919 engagements

"1/ In crypto its dont trust verify. In AI its well we still mostly trust"
X Link @inference_labs 2025-10-11T16:00Z 25.7K followers, 3162 engagements

"3/ The result: modular distributed and practical zkML for real-world workloads. A new era of efficient targeted proofs that scale with you. Discord: TG: Whitepaper: Repo:"
X Link @inference_labs 2025-10-09T16:00Z 25.5K followers, 1261 engagements

"1/ Crypto is racing ahead with RWAs DePIN and $126K BTC. AI is exploding with agentic systems. The convergence is inevitable"
X Link @inference_labs 2025-10-11T00:00Z 25.7K followers, 2735 engagements

"Prove you're XX without sharing your ID. Inference Labs' Age Verification model utilizes on-device AI and zkML to verify age without exposing private data. Built on Subnet X it's live open-source and regulation-ready. Privacy verified"
X Link @inference_labs 2025-06-10T15:00Z 25.6K followers, 27.9K engagements

"Huge thanks to @KeithSingery for diving deep into DSperse Subnet X and our vision for the future of zkML. Its not just about proof of inference its about making AI verifiable scalable and real-world ready. Watch below to learn how DSperse unlocks a new dimension of scalability for Subnet X and beyond"
X Link @inference_labs 2025-10-16T14:24Z 25.6K followers, 3780 engagements

"Why prove everything when only some layers impact the decision DSperse = lighter circuits faster proofs scalable zkML"
X Link @inference_labs 2025-10-12T00:00Z 25.7K followers, 3231 engagements

"2/ What if platforms could verify your age without ever seeing your face or storing your ID With on-device AI and zero-knowledge proofs Colin explains how Subnet X is prototyping exactly that a future where privacy and compliance dont conflict"
X Link @inference_labs 2025-06-09T18:00Z 25.6K followers, 4007 engagements

"1/ Tired of bulky proofs in zero-knowledge ML DSperse slices your models into logical segments letting you verify just the high-impact operations faster lighter and with less friction"
X Link @inference_labs 2025-10-09T16:00Z 25.6K followers, 29.7K engagements

"Inference Labs x Unmarshal: Were thrilled to team up with @UnmarshalAI a powerful multi-chain data network serving real-time AI-enhanced insights across 55+ blockchains. Together were bringing trustless AI verification to verifiable on-chain data powering the next wave of DeFi RWA DePIN and AI-native apps with clarity and integrity. In Web3 access isnt enough. You need proof"
X Link @inference_labs 2025-07-14T14:38Z 25.6K followers, 22.6K engagements

"AI agents are exploding in 2025 from logistics to finance. But XX% are expected to fail due to trust gaps. Were fixing that with cryptographic proofs for every action. Join the movement:"
X Link @inference_labs 2025-10-09T00:00Z 25.6K followers, 3138 engagements

"DSperse slices the lifecycle into modular pieces compile prove verify orchestrated for speed. The result: PK/VK sizes XX% smaller proof times cut by 60%"
X Link @inference_labs 2025-10-10T16:00Z 25.6K followers, 3858 engagements

"2/ At Inference Labs were fixing that with zk-proofs for AI outputs. Be part of the trust revolution. Join Discord:"
X Link @inference_labs 2025-10-11T16:00Z 25.6K followers, 1580 engagements

"2/ Inference Network sits at that intersection: zkML proofs that make AI outputs auditable portable and trustworthy. AI x Web3 isnt hype its infrastructure"
X Link @inference_labs 2025-10-11T00:00Z 25.6K followers, 1290 engagements

"Everyone wants AI you can trust. The default answer: prove everything. Wrap the model in circuits attach zk proofs done. Problem Its slow burns memory and never leaves the lab. Real systems need receipts only for high-risk parts"
X Link @inference_labs 2025-10-13T00:00Z 25.6K followers, 3188 engagements

"Online age checks are broken either invasive or easy to fake. Inference Labs enables zk-based age verification that proves eligibility without exposing personal data. HIPAA-compliant. On-chain ready. Privacy-first infra live on Subnet 2"
X Link @inference_labs 2025-07-10T14:30Z 25.6K followers, 1231 engagements

"1/ In collaboration with @ComputeHorde Subnet X became the first to adopt Yuma3 a major step forward in reducing the costly impact of weight copying across the network"
X Link @inference_labs 2025-08-14T17:20Z 25.6K followers, 26.3K engagements

"AI agents are transforming the landscape across various sectors. But without verification its just hype. Inference Labs powers Bittensor Subnet X with the largest zkML proving cluster delivering Proof-of-Inference at scale. This is math over marketing"
X Link @inference_labs 2025-06-07T14:00Z 25.6K followers, 26.5K engagements

"1/ A recent conversation with our co-founder Colin sheds light on how Inference Labs is tackling one of the hardest problems in AI x regulation: privacy-preserving KYC"
X Link @inference_labs 2025-06-09T18:00Z 25.6K followers, 22.7K engagements

"Why prove the whole model when strategic slices deliver trust where it matters DSperse targets high-value subcomputations with up to XX% smaller circuit files and XX% faster proofs on real models. Modular zkML starts here"
X Link @inference_labs 2025-10-04T16:00Z 25.5K followers, 3929 engagements

"Like most computer nerds we're camera shybut we appreciate the platform and the support Heres a video of our own @colingagich for the culture"
X Link @inference_labs 2024-10-18T15:19Z 25.6K followers, 183.6K engagements

"Most AI infra wasnt built to be trusted. Ours is. Inference Labs provides the backbone for verifiable AI cryptographic proof that your model ran honestly at speed and scale. Weve built the largest zkML proving cluster on Bittensor Subnet X processing 10M+ proofs/month with up to XX% faster performance than leading alternatives. Our stack includes: Model-to-circuit conversion tooling Scalable proof operations (24/7) Open-source zkML contributions adopted across DeAI Now were building a sovereign network for verifiable inference designed for performance privacy and trust at the protocol layer."
X Link @inference_labs 2025-07-13T15:45Z 25.6K followers, 23.3K engagements

"AI without proof is a trust crisis waiting to happen. Inference Labs powers the largest zkML proving cluster on Bittensor (Subnet 2) making AI outputs verifiable by design like rollups but for inference"
X Link @inference_labs 2025-08-04T16:00Z 25.6K followers, 23.4K engagements

"AI that cant prove its output is a liability. Inference Labs converts models into zk-circuits and delivers cryptographic proof in real time XX% faster than others. Already powering 10M+ proofs/month on Subnet X. This is trust at compute speed"
X Link @inference_labs 2025-07-31T16:00Z 25.6K followers, 2051 engagements

"1/ AI without proof is a trust crisis waiting to happen. When models can say anything and no one can verify what ran or how chaos is only a prompt away"
X Link @inference_labs 2025-08-18T16:10Z 25.6K followers, 23.9K engagements

"2/ Inference Labs powers the largest zkML proving cluster on Bittensor (Subnet 2) making AI outputs verifiable by design. Think rollups for inference: provable decentralized auditable"
X Link @inference_labs 2025-08-18T16:10Z 25.6K followers, 1213 engagements

"2/ Our zkML stack ensures: The right model was used The logic wasnt modified post-deploy The output can be publicly audited without leaking inputs This isnt just for compliance its for operational resilience"
X Link @inference_labs 2025-07-21T15:00Z 25.6K followers, 1599 engagements

"3/ NET model: - Total proof time: drops by XXXX% - Compile phase: XXXX% faster - Prove phase: XXXX% faster - Verify phase: XXX% slower - PK size: XX% smaller - VK size: XX% smaller"
X Link @inference_labs 2025-10-09T16:00Z 25.5K followers, 1476 engagements

"🌍 Over XX% of consumers are concerned about sharing their personal data online. Inference Labs enables zero-knowledge age & ID verification no docs no data leaks. Private proofs. Public trust. This is zk identity live on Subnet 2"
X Link @inference_labs 2025-07-02T17:00Z 25.6K followers, 1518 engagements

"1/ Ever wondered why zkML hasnt scaled Proving entire models is slow costly and impractical. What if you could prove only the parts that matter"
X Link @inference_labs 2025-09-29T16:00Z 25.6K followers, 21K engagements

"Smart contracts gave us trustless money. Inference Labs gives us trustless AI powering Bittensor Subnet X with the largest zkML cluster. Proof of Inference ensures agents dont just claim outputs they prove them cryptographically"
X Link @inference_labs 2025-06-13T14:45Z 25.6K followers, 22.4K engagements

"2/ Inference Labs adds cryptographic accountability to every AI decision with 10M+ proofs/month on Subnet 2"
X Link @inference_labs 2025-08-03T15:45Z 25.6K followers, 1340 engagements

"1/ Autonomous systems are moving from science fiction to your daily workflow. Self-driving cars. AI agents. Decentralized robots"
X Link @inference_labs 2025-10-13T16:00Z 25.7K followers, 3078 engagements

"Agentic AI is surging but trust is lagging. XX% of companies are experimenting with AI agents and Gartner calls it the fastest-rising tech of 2025. At Inference Labs were building the missing layer: cryptographic proofs for every action an agent takes"
X Link @inference_labs 2025-10-08T00:00Z 25.7K followers, 3580 engagements

"3/ Deployed on Bittensor Subnet X processing 10M+ zk-proofs/month. Inference Labs is turning verifiability into infrastructure. Because when AI fails in the real world proof beats hindsight"
X Link @inference_labs 2025-07-21T15:00Z 25.6K followers, 1532 engagements

"Full-model circuitization forces compromises. DSperse lets devs pick the pipelines slices and level of audit preserving fidelity privacy and speed. ML trust on your terms"
X Link @inference_labs 2025-10-06T16:00Z 25.6K followers, 4299 engagements

"Full-circuit zkML proofs are heavy large proving keys slow proofs and high memory cost. With DSperse slicing the Doom model shows: - Proof time cut from 200s 145s - PK size shrinks by 50%"
X Link @inference_labs 2025-10-02T16:00Z 25.7K followers, 3376 engagements

"Inference costs are down 280x since 2022. AI is cheap. AI is fast. But is AI verifiable Thats where we come in. Join our socials to see what verifiable AI looks like. TG: Discord:"
X Link @inference_labs 2025-10-16T00:00Z 25.6K followers, 2713 engagements

"Think modular rollups but for machine learning proofs. Dsperse breaks down zkML into slices: lighter compile faster proving smaller keys. The payoff Proof generation XX% faster and half the memory footprint. Scaling verifiable AI without the bottlenecks"
X Link @inference_labs 2025-10-12T16:00Z 25.7K followers, 3591 engagements

"Slicing isnt just faster it enables modular distributed and scalable zkML. On the NET model: - Proof time drops from 229s 91s - PK size reduced from 8.6GB 2.8GB - VK size: 1172 vs 2434 unsliced Enables distributed computation: parallelize proofs for larger models"
X Link @inference_labs 2025-10-05T16:00Z 25.6K followers, 5984 engagements