Providing info on Ordinals. Theorizing on Digital Matters

Layer 23 retweeted
Trac just shipped a full RTS game where your opponent is a fully autonomous AI agent. No wallet. No tokens. Free to play. Here's the part that actually matters. The agent pays real compute providers in real time to run every match. No company can sit in that loop and take a cut or approve the transaction first. That only works with P2P rails. - Stripe needs a merchant account - A cloud provider needs a company behind it - An autonomous agent negotiating with a stranger's GPU mid match needs none of that No central server sits between the agent and the provider. They find each other directly and settle it themselves. This is the actual blueprint for agent to agent commerce. Not a whitepaper case. A video game running on it right now. Go see if you can beat it.
Moving Out! - Enjoy "Wüste Zwei", a real-time strategy game based on a popular sci-fi book about desert planets, worms, and competing houses... Relive the full experience with 3 campaigns, 27 maps and skirmish mode - it's more fun than ever: openmayhem.ai/game/ More info below. This version is a faithful recreation of the classic PC/Amiga game, which founded the actual genre. But this time you fight agentic-controlled opponents. It demonstrates how Trac Network and OpenMayhem enable real-world usage of AI & modern Crypto. No player needs any wallet, tokens, or to spend anything to play. All that while the underlying economy works. While it works on phone, Desktop is preferred. In-game, hit the ESC key for how-to and options. F-key enables fullscreen. Under the hood: Your opponent is a fully autonomous agent using OpenMayhem providers. We also used OpenMayhem providers to create media assets for the game itself (videos, images, voice narration). The game engine was created using Fable, using its charming art style. You can play the game fully for free, we sponsor the OpenMayhem provider fees. The game is the top-facing layer of what we are building: Trac Network -> OpenMayhem -> Wüste Zwei If you think it's cool, we'll be willing to open-source and update it further! Feel free to report bugs. The network is still young and has limited ai providers. So if you encounter "uplink" problems in-game (see top right), then help the network and join as provider. The more providers, the more can play. As a bonus, you can earn being a provider.
1
10
32
1,955
Layer 23 retweeted
Only a matter of time until this project becomes a top 10 imo trac-network:native
Below monthly net. earning estimates for providers w/ consumer hardware, serving models on OpenMayhem. It assumes 24/7 use for model qwen3.6-35b-a3b (very popular) on seed pricing. Effective earnings depend on uptime, demand & max. context.
3
14
49
1,938
And there you have it! After Venice.ai, @TracNetwork now provides a nice alternative but in a more decentralized way, bringing LLM onto the P2P rail. The trend now is cheap and decentralized AI anyway. Big AI companies struggle to bring costs down
OpenMayhem 0.2.1 released. With Ace Step 1.5 support. Providers can now serve & earn from music generation and users/agents can buy. Quick sample below.
1
4
18
349
Layer 23 retweeted
Video generation soon
OpenMayhem 0.2.1 released. With Ace Step 1.5 support. Providers can now serve & earn from music generation and users/agents can buy. Quick sample below.
2
15
49
1,526
Layer 23 retweeted
Introducing OpenMayhem, Trac's flagship open-source inference router, inviting everyone to provide & use compute. All links at the end of this post. Open-weight models are not only getting smarter but fit on consumer hardware. Hardware that you already own. The increasing use of open-weight models for agentic loops demand for an open market to cost-effectively choose the best models for the job. Even if a model fits, it competes with harnesses & other memory-consuming processes. Clouds are increasingly becoming the bottleneck: participation in existing markets are gate-kept, adding additional costs for inference. Almost everything that happens is opaque. Agents have to go through all sorts of hoops just to access inference instead of auto-explore and use. Most of the hardware capable of solving this problem are laying dormant, forcing everything & everyone into data centers. We took all this into account and built OpenMayhem on top of a Trac Network subnet: - Direct & fast p2p routing between agents & inference providers (you) - Open, automated inference markets to explore & access the best prices - Extensive hardware support, from ordinary computers/laptops, to enthusiast setups & pro-hardware - Publicly reproducible via Trac Network (oracle & evidence ledger) - Multiple payment rails - Everyone can earn & use, permissionless We are in the process of rolling out many more supported models for OpenMayhem's inference providers. If you need specific ones, feel free to let us know. Explore OpenMayhem 🌐 Try OpenMayhem: openmayhem.ai 📖 Read the full launch article: medium.com/p/6e73c978a3f0 💻 Explore and install the open-source software: github.com/Trac-Systems/open… 🔗 Learn about the network beneath it: tracsystems.io/trac-network
15
50
95
225,079
Layer 23 retweeted
OpenMayhem is using the full stack, end-to-end: Trac Network subnet <=> p2p <=> contracts <=> inference. An open-source automated markets for inference, users & providers.
Introducing OpenMayhem, Trac's flagship open-source inference router, inviting everyone to provide & use compute. All links at the end of this post. Open-weight models are not only getting smarter but fit on consumer hardware. Hardware that you already own. The increasing use of open-weight models for agentic loops demand for an open market to cost-effectively choose the best models for the job. Even if a model fits, it competes with harnesses & other memory-consuming processes. Clouds are increasingly becoming the bottleneck: participation in existing markets are gate-kept, adding additional costs for inference. Almost everything that happens is opaque. Agents have to go through all sorts of hoops just to access inference instead of auto-explore and use. Most of the hardware capable of solving this problem are laying dormant, forcing everything & everyone into data centers. We took all this into account and built OpenMayhem on top of a Trac Network subnet: - Direct & fast p2p routing between agents & inference providers (you) - Open, automated inference markets to explore & access the best prices - Extensive hardware support, from ordinary computers/laptops, to enthusiast setups & pro-hardware - Publicly reproducible via Trac Network (oracle & evidence ledger) - Multiple payment rails - Everyone can earn & use, permissionless We are in the process of rolling out many more supported models for OpenMayhem's inference providers. If you need specific ones, feel free to let us know. Explore OpenMayhem 🌐 Try OpenMayhem: openmayhem.ai 📖 Read the full launch article: medium.com/p/6e73c978a3f0 💻 Explore and install the open-source software: github.com/Trac-Systems/open… 🔗 Learn about the network beneath it: tracsystems.io/trac-network
4
14
39
1,987
Layer 23 retweeted
We are right now rolling out the launch roster with OpenMayhem inference router support for essential models. All listed models below run & serve on consumer hardware. Once they serve, users/agents can immediately quote & use them. Price floors below are initial & will automatically adjust based on demand. Heavier models like GLM-5.2 will be added after the full launch roster is live. Providers can already choose which ones to serve from the live ones. Live now: - Qwen3.6-35B-A3B Uncensored — $0.05/1M tokens - Gemma-4-E4B — $0.015 in / $0.05 out per 1M tokens - Z-Image-Turbo — $0.0025/image - Parakeet-TDT-0.6B-v3 (STT) — $0.001/audio minute Next in line: - ACE-Step 1.5 (music) — $0.0001/audio second - needle (tool-calling) — $0.002 in / $0.008 out per 1M tokens - Huihui-Agents-A1 Uncensored (vision) — $0.05/1M tokens - Ternary-Bonsai-27B — $0.03 in / $0.18 out per 1M tokens Remaining: - gpt-oss-20b — $0.0125 in / $0.05 out per 1M tokens - gpt-oss-120b — $0.02 in / $0.10 out per 1M tokens - Qwen3.6-35B-A3B — $0.055 in / $0.40 out per 1M tokens - Ornith-1.0-9B — $0.025 in / $0.10 out per 1M tokens - Devstral-Small-2-24B (coding) — $0.03 in / $0.08 out per 1M tokens - Qwen3.5-9B Uncensored — $0.03/1M tokens - Qwythos-9B MTP — $0.025 in / $0.10 out per 1M tokens - FLUX.2-klein-4B — $0.0025/image - Qwen-Image-2512 — $0.0075/image - Chroma1-HD — $0.005/image - Kokoro-82M (TTS) — $1.50/1M characters - Chatterbox (voice clone TTS) — $5.00/1M characters - whisper-large-v3-turbo (STT) — $0.0015/audio minute - bge-m3 / Qwen3-Embedding-0.6B — $0.004/1M tokens - bge-reranker-v2-m3 — $0.01/1M tokens
5
23
79
14,386
Achieved using 4X smaller local models. Curious what happens when you synthwave frontier models 🤔
Synthwave is now officially in frontier model territory. We ran the latest AIME 2026 benchmark, a set of hard math tasks, against Synthwave using an ensemble of 4 small local models across 2 DGX Spark machines. It outperformed DeepSeek V3.2, putting it at rank #4. None of the involved models would achieve this alone. The interesting part is that this scales: it can uplift even smaller models running on single machines, or combine the most powerful frontier models into an even stronger intelligence layer. For details check out the project: github.com/Trac-Systems/synt…
3
304
Layer 23 retweeted
Most staking: Lock tokens. Earn inflation. Hope price goes up. TrustLink staking: Lock TNK. Resolve real disputes. Earn from real trading volume. Yield backed by actual economic activity. Not printer go brrr. 👉 trustlinkmarket.com → ⚖ Arbitrate @TracNetwork $TNK
6
16
339
Layer 23 retweeted
There are two types of crypto builders: Ones who write whitepapers about P2P escrow. Ones who ship it. TrustLink shipped. 12,500 TNK. 1,055 users. 13% of @TracNetwork. 👉 trustlinkmarket.com $TNK #BuildInPublic
6
24
290
Layer 23 retweeted
𝐓𝐀𝐏 𝐂𝐫𝐨𝐬𝐬𝐃𝐄𝐗 𝐂𝐡𝐚𝐢𝐧 #𝟐 ⛓️ @solana Solana users, meet Bitcoin 🤝 From June 7th, all Solana assets will be swappable for any on TAP Protocol (and vice versa). Permissionless and fully self-custodial cross-chain swaps on Bitcoin, only with TAP CrossDEX 🔁
25
55
148
10,965
Layer 23 retweeted
Numbers don't lie: 13%+ of Trac Network ✅ 12,500 TNK volume ✅ 1,055 new users onboarded ✅ Built in 4 weeks. Zero marketing. Still growing. This is what product-market fit looks like. 👉 trustlinkmarket.com @TracNetwork $TNK
7
20
312
Layer 23 retweeted
TAP Wallet now supports HTLCs 🔐 For the first time, an Ordinals-native wallet supports Hashed Timelock Contracts — a major step forward for Bitcoin-native utility HTLCs unlock more advanced transaction logic and lay important groundwork for swaps with greater trust-minimisation
18
47
141
8,719
Layer 23 retweeted
Dare: Find a P2P platform that: ✅ Works on mainnet ✅ Handles BTC, ETH, DOGE, Ordinals ✅ No KYC ✅ Staked arbitrators ✅ Can't be shut down We'll wait. 👉 trustlinkmarket.com @TracNetwork $TNK
6
25
385
Layer 23 retweeted
Arbitrating on TrustLink is the closest thing to a decentralized job. Stake TNK. Review evidence. Rule fairly. Get paid. No employer. No office. No CV required. Just your stake and your judgment. 👉 trustlinkmarket.com → ⚖ Arbitrate @TracNetwork $TNK
8
19
581
Layer 23 retweeted
Can’t wait for Bitcoin Native Perps Market!!!
The TAP Protocol upgrade has been performed will activate at block 952317 Everyone running Ord-TAP will need to perform an update once the upgrade goes live. TAP Reader users don't need to do anything. Switch on notifications 🔔 As you're going to want to see what we do next 👀 Read more below ⬇️
3
15
371
Layer 23 retweeted
Basically the only serious player that still builds on Bitcoin (and more via Trac). Long term vision hasn't changed. Miners will love activities on Bitcoin more and more as halvings happen.
The TAP Protocol upgrade has been performed will activate at block 952317 Everyone running Ord-TAP will need to perform an update once the upgrade goes live. TAP Reader users don't need to do anything. Switch on notifications 🔔 As you're going to want to see what we do next 👀 Read more below ⬇️
2
9
36
1,715