Pinned
Building open-source LLMs & benchmark for crypto, DeFi and RWAs.
Our benchmark hit #1 on @huggingface datasets.
- Introducing DMind-1: an open-source large language model purpose-built for Web3 scenarios. DMind-1(32B) is fine-tuned on Qwen3-32B by @Alibaba_Qwen, using tons of Web3-specific knowledge, purposefully optimized to balance performance and cost for AI+Web3 applications. DMind-1
- Introducing DMind Benchmark: the first domain-specific benchmark to evaluate LLMs in Web3. Features: ✅ Covers 9 core categories in #Web3— Blockchain Fundamentals, Infrastructures, Smart Contracts, DeFi, DAO, NFT, Token Economics, Meme, and Security. ✅ 1,917 questions —
- DMind’s goal is to develop, iterate, and evolve together with the community. LLMs shouldn’t be exclusive to tech giants. We're committed to building a truly open AI.
- Cooking Contributor Program... We’re building LLMs in a way you’ve never seen before. A way that everyone can take part in.
- Introducing DMind Benchmark: the first domain-specific benchmark to evaluate LLMs in Web3. Features: ✅ Covers 9 core categories in #Web3— Blockchain Fundamentals, Infrastructures, Smart Contracts, DeFi, DAO, NFT, Token Economics, Meme, and Security. ✅ 1,917 questions —
- Introducing DMind Benchmark: the first domain-specific benchmark to evaluate LLMs in Web3. Features: ✅ Covers 9 core categories in #Web3— Blockchain Fundamentals, Infrastructures, Smart Contracts, DeFi, DAO, NFT, Token Economics, Meme, and Security. ✅ 1,917 questions —
- 🚀 We're proud to join the next wave of Sentient Data Consortium Partners! At @DMind_ai, we’re building the crypto knowledge layer for open-source AGI and Web3 expert LLMs — through rigorous evaluation, high-quality dataset, and open contribution. As a data partner, we’re🚀ANNOUNCING THE NEXT WAVE OF SENTIENT DATA CONSORTIUM PARTNERS Sentient is building the largest collection of intelligence in the world, enriched with the best agents, models, and domain-specific data. Today, we’re thrilled to welcome another wave of partners whose
- Replying to @dmind_aiRound 2: Describe a realistic delta-neutral approach that earns from funding payments.
00:00 - Replying to @dmind_aiLet’s see the battle between DMind-1($0.3/MT) and GPT-o3($10/MT). Round 1: What’s the difference between MakerDAO and Aave, and which is better for borrowing stablecoins with ETH?
00:00 - Replying to @SentientAGIWe’re thrilled to officially join the Sentient Data Consortium! 🚀 At DMind.ai, we believe that open, verifiable knowledge is the foundation for truly trustworthy agents. By combining our benchmarked Web3 models and transparent data with Sentient’s mission, we’re
- DMind 🤝 Sentient We're thrilled to join the Sentient Data Consortium! With open benchmarks, Web3 expert models, and transparent data, we're pushing the boundaries of crypto-native AGI together. Let’s build the brain of open-source intelligence for Digital finance!
- Replying to @dmind_aiWe used the DMind Benchmark to test DMind-1 and other flagship LLMs on Web3 tasks across 9 key areas: Blockchain Fundamentals, Infrastructure, Smart Contract, DeFi, DAO, NFT, Tokenomics, Meme, and Security. DMind-1 matches top-tier models on core Web3 knowledge at a much lower













