ModelOS is live! Point your rigs at our pool and start mining $MDL. Merge-mined with $PRL for double rewards. → Mine: pool.modeloslab.xyz/mine → Discord: discord.gg/VawUAUHQ3 → Try inference models: compute.modeloslab.xyz/ A milestone for decentralized inference. Thanks for being here early.

dCloud is going live today. We invite everyone to explore what it is, how it fits into the ModelOS ecosystem, and why it matters for the future of blockchain-based AI compute.

$MDL is now listed on SafeTrade. A new step for ModelOS as we continue building a Proof-of-Useful-Work Layer 1 for GPU mining, AI compute, dCloud, and future inference utility. safetrade.com/exchange/MDL-U…

We're now on all the biggest Pearl pools, merge-mining MDL with pearl-2:native. Network hashrate just hit 22 EH/s, which makes Modelos extremely stable and secure. Now we're building the next big thing: a plugin that turns idle compute into a decentralized inference and training marketplace. In testing now: → Inference hosting: run any open-source model on your hardware, serve requests, get paid → GPU pool rental: contribute spare capacity to a shared pool, earn based on usage → Shard hosting: not enough VRAM for a full model? Host a piece of it alongside others → Decentralized training: contribute compute toward training a shared model, collaboratively, from genesis → The first fully community-trained model, no centralized infra, just distributed contributors Mining was step one. This is step two.

OTC is live. $MDL is now tradable. Pay in: → USDC or USDT on Base, Arbitrum, Ethereum, BNB, Solana → BTC → Pearl ($PRL) otc.modeloslab.xyz/

Found a very early project, right in the most bullish narrative right now. An L1, a fork that improves on @prlnet, built by @ModelosMDL. Here's the part that makes it different: Modelos turns mining into real AI work. Run the vLLM miner and your GPU is serving actual model inference while it mines, the same compute that secures the chain answers real prompts. The work isn't wasted hashing, it's useful output someone can pay for. Why it's interesting if you mine: 🔹 Merge-mined with Pearl. Already mining $PRL on a GPU? Point the same card at $MDL and mine both at once. Same hashpower, no extra cost, second coin. 🔹 Two incomes from one GPU. Block rewards plus inference fees for serving models. Miss a block, your card still earns answering prompts. 🔹 Fair launch, no team allocation, no VC bags. 21M cap, Bitcoin's number on purpose. Heading in discord that several major Pearl pools are already planning to add dual mining with MDL, so distribution is about to widen fast. Network just crossed 200 PH/s and it's still early. Not trading yet, but OTC is coming soon. Site: modeloslab.xyz/ Whitepaper: modeloslab.xyz/modelOS_White… Discord: discord.gg/VawUAUHQ3 $MDL

Most crypto mining is simple: burn electricity for one reward. MDL changes that. With one GPU, you can: ✅ Merge-mine Pearl and MDL at the same time. ✅ Earn two streams of rewards instead of one. ✅ Support a network focused on decentralized AI inference—not just another speculative token. What I like most is the launch approach. No VC allocation. No massive insider unlocks. No expensive ASIC advantage. Just GPUs securing the network while helping build infrastructure for AI workloads. As AI demand keeps growing, compute becomes one of the most valuable resources in crypto. Instead of letting GPUs sit idle, MDL puts that hardware to work while still allowing Pearl mining through merge mining. If you’re already mining Pearl, it’s worth asking yourself: Why settle for one reward when your GPU can earn two? That’s a much more interesting future than another copy-paste mining project. @ModelosMDL #MDL #GPUmining #CryptoMining #AI #Pearl

НОВАЯ МОНЕТА ModelOS (MDL) — Майним Одновременно с Pearl! piped.video/MQV5ABTiRJQ?si=cYz3… через @YouTube

This is exactly the kind of walkthrough new miners need to mine MDL. Huge thanks to @RedPandaMining for putting this together. If you’ve got a GPU and you’re curious how Modelos works, start here.

👀👀👀

New milestone: the ModelOS network just crossed 200 PH/s We've had a lot of interest from miners looking to leave PRL-only pools to dual-mine MDL + PRL, so expect this to keep climbing over the next few days. Still early. What's coming: → OTC website, real soon → X content contest just opened. 1000 MDL in prizes. Details in our Discord: discord.gg/VawUAUHQ3 → Always open to collabs, PRL pools, OTC, CEX. DMs open → A campaign to test our decentralized inference We're just getting started.

About 72 hours since mainnet launched, and things are unfolding exactly as planned. ASERT update: it’s activated and playing out exactly as expected. We’re nearly at block 17,000, and by block 17,500 block times will settle to a steady ~194s, matching Pearl. The network’s finding its rhythm right on schedule. Hashrate has held strong the whole way, consistently around 60-80 PH/s, with plenty of stretches above 100 PH/s. And we’re just getting started. Next, we’d love to help @prlnet PRL pools add $MDL. Your miners would earn it at no extra cost: same hashpower, a second coin. Any pool, DM us, we’re always happy to collaborate. After that: OTC opens, and from there we get inference going, priced lower than what comparable models cost today. We’re still early, and it’s looking good.

We hit some hiccups in the first hours after launch. When hashrate ramped up far faster than anything we saw in testing, it surfaced issues that never came up before. The dev team worked around the clock and we've fixed most of them. The main pool is stable now, we added cores to the explorers so other peers can connect, and we sorted out HiveOS. Still heads-down on the rest. Next up: → Testing inference, making sure miners on the vLLM miner earn their inference rewards → Waiting to hit our first Pearl block → Adding OTC, then CEX. More as we go. Blocks still very short right now, so it's a good time to get your rigs on the network.

We're later than some wanted, but look what's ready: → Website live: modeloslab.xyz/ → A huge lineup of open models served by miners, our MoE already running on our own miner in testing. More models being added after testing → Image generation working (cursed two-eyed cat below as proof) → Mining bugs fixed, pool now merge-mines with Pearl cleanly with strong hashrate on every miner we tested → API inference in testing, live at launch. Use it with Codex, OpenClaw, whatever you run → Major breakthrough in testing: a decentralized GPU shard pool, opening the door to multiple miners hosting a single model collaboratively An insane number of models to play with, decentralized inference at the core. Enjoy your weekend. Launch is (very) early next week.

The whitepaper is here. 📄 modeloslab.xyz/modelOS_White… No teasers, no fluff. The full picture of what ModelOS is and how it works. Pearl merge-mining, the dual $PRL / $MDL pool, and on-chain verifiable inference. Test is rounding up too, we're on cleanup. Home stretch. This is a fair launch. No special access, no head start for anyone. Just be ready when it goes. Hang with us.