Dmind
DMind is an open-source AGI research organization developing large language models and evaluation benchmarks specifically for Web3 and digital finance, serving developers, funds, protocols, and auditors through community-driven distribution on Hugging Face, GitHub, and Discord.
- Company typePrivate
- Founded2025
- Headquarters—
- Headcount11–50
- GTM typeB2B
- OfferingSoftware
What Dmind does
DMind is an open-source AGI research organization that develops large language models and evaluation infrastructure targeted exclusively at Web3 and digital finance use cases. Founded as a private entity operating with 11–50 employees, the company serves Web3 developers, funds and protocols, auditors, and advanced agent ecosystems that require AI assistance for security-critical, irreversible financial execution in decentralized markets.
The company's core product is the DMind-3 series, a sovereign Edge-Local-Cloud intelligence stack composed of three cooperating layers: Nano acts as a deterministic intent firewall at the edge to secure present transactions, Mini operates as a private high-fidelity reasoning engine on user hardware to formulate immediate strategy, and the DMind-3 cloud tier serves as a policy-governed global context synthesizer that uses policy-driven selective offloading to route computation by privacy sensitivity and uncertainty. DMind also maintains prior-generation models (DMind-1, DMind-2) and an evaluation suite comprising the DMind Benchmark (3,543 expert questions across 9 Web3 domains, evaluating 31 leading AI models including GPT-5, Claude, and Gemini), FinanceQA, and AIME 2025. The DMind Benchmark was accepted to the KDD 2026 Dataset & Benchmark Track and reached #1 trending status on Hugging Face upon release in April 2025, with over 9,650 downloads accumulated by January 2026.
DMind operates a community-led, product-led go-to-market: all products are released under permissive open-source licenses and distributed through Hugging Face, GitHub, and Discord, with academic visibility via KDD 2026 as a positioning channel. The company has not disclosed a direct pricing model or revenue stream; monetization, if any, is indirect via third-party API providers OpenRouter (unified model API) and NanoGPT (anonymous pay-per-prompt in crypto or card). Strategic relationships include academic collaborations with Zhejiang University and Nanyang Technological University for benchmark research and a partnership with Minara to translate benchmark findings into production Web3 tools. No funding rounds, investors, ownership structure, or revenue figures have been disclosed.
Dmind firmographics
Firmographics- Name
- Dmind
- Legal name
- DMind
- Website
- https://dmind.ai
- Company type
- Private
- Founded year
- 2025
- Operating status
- Operating
- Headcount range
- 11–50 employees
- Short description
- DMind is an open-source AGI research organization developing large language models and evaluation benchmarks specifically for Web3 and digital finance, serving developers, funds, protocols, and auditors through community-driven distribution on Hugging Face, GitHub, and Discord.
- Ownership category
- akta.pro rank
Dmind industry classification
Industry- Product category
- AI Foundation Models for Web3 and Digital Finance
- NAICS
- Software Publishers (5132)
- SIC
- Services-Prepackaged Software (7372)
- akta.pro primary industry
- On-Device/Edge Foundation Models (Mobile/Embedded LLMs) (HDAAACAN)
- akta.pro secondary industry
- Smart Contract Frameworks & SDKs (developer kits, templates, standards) (FSAPABAB)
Keywords
Dmind business model
Business model- GTM type
- B2B
- Offering type
- Software
- Cost components
- Technology or R&D, Personnel, Infrastructure, Marketing or Sales, Operations
Revenue model
- Open-source products and research: DMind releases products under permissive open-source licenses. Revenue generation mechanisms are not explicitly disclosed in available sources.
Pricing tiers
| Model | Billing | Price |
|---|---|---|
| Freemium | Others | Open-source free tier |
Go-to-market motion2 records
Distribution channels4 records
Marketing channels5 records
Dmind product offering
Product offeringCore offering
DMind is an open-source AGI research organization that develops and releases large language models (the DMind-1, DMind-2, and DMind-3 series) and AI benchmarks purpose-built for digital finance and Web3. The flagship DMind-3 stack combines an edge intent firewall (Nano), an on-device reasoning engine (Mini), and a policy-governed cloud tier to secure irreversible financial execution against adversarial risks. All products are released under permissive open-source licenses, with API access routed through partners such as NanoGPT and OpenRouter.
Product overview
DMind is an open-source AGI research organization building products for next-generation digital finance. The product portfolio consists of: (1) the DMind-3 series — a sovereign Edge-Local-Cloud intelligence stack with three cooperating layers (Nano as edge intent firewall, Mini as private reasoning engine, and DMind-3 cloud tier as context synthesizer); (2) prior generation LLMs (DMind-2 and DMind-1); (3) evaluation benchmarks including DMind Benchmark (for Web3 AI evaluation, accepted at KDD 2026), FinanceQA, and AIME 2025; and (4) API access products through NanoGPT and OpenRouter for model access. All products are open-source and released under permissive licenses.
Differentiator
Problem solved
Functional benefit
Brands
- DMind-3: Sovereign Edge-Local-Cloud intelligence stack designed to secure irreversible financial execution in Web3 environments, consisting of Nano (edge intent firewall), Mini (private reasoning engine), and DMind-3 cloud tier.
- DMind Benchmark
- FinanceQA
- AIME 2025
Products and services
- DMind-3 LLM (cloud tier) Policy-governed global context synthesizer and the cloud tier of the DMind-3 sovereign Edge-Local-Cloud intelligence stack, designed to secure irreversible financial execution in Web3 environments by routing computation based on privacy sensitivity and uncertainty without exporting private intent by default.
- DMind-3 Nano Edge-tier deterministic intent firewall in the DMind-3 sovereign stack; secures the present Web3 transaction at the edge against adversarial execution.
- DMind-3 Mini Private high-fidelity reasoning engine that runs on user hardware to formulate immediate strategy for Web3 transactions; the on-device tier of the DMind-3 sovereign stack.
- DMind Benchmark Open-source evaluation benchmark consisting of 3,543 expert questions across nine Web3 domains, used to evaluate AI systems (31 leading models including GPT-5, Claude, and Gemini tested) on security vulnerability detection, token economics reasoning, and related safety-critical tasks; accepted to the KDD 2026 Dataset & Benchmark Track.
- FinanceQA Open-source benchmark dataset for evaluating AI model performance on financial question-answering tasks.
- AIME 2025 Open-source benchmark for evaluating AI model performance on mathematical reasoning tasks.
Quantifiable outcome
- No AI model is production-ready for safety-critical Web3 tasks according to DMind Benchmark findings
- +1 more outcomes
Companies that use Dmind
Customer profileSegments4 records
Ideal customer profiles2 records
Dmind technology and API
TechnologyTechnology focussed Yes
API detail
- Has API
- Yes
- API docs
- API detail
Core technology
AI maturity
App detail
Integration2 records
AI capability8 records
Feature4 records
Dmind partnerships and signals
Strategic signalPartnerships
Three partnerships are on record, tiered core.
- MinaracoreStrategic partnership with Minara to translate DMind Benchmark research findings into real-world AI tools for Web3 users, auditors, and developers. Minara contributes as a community contributor with logo displayed on DMind's contributors page.
- Zhejiang UniversitycoreAcademic research collaboration for the DMind Benchmark paper, contributing researchers to evaluate AI models for Web3 applications.
- Nanyang Technological UniversitycoreAcademic research collaboration for the DMind Benchmark paper, contributing researchers to evaluate AI models for Web3 applications.
Scale indicators4 records
Recent moves6 records
Expansion highlights4 records
Dmind competitors and assessment
Company assessmentBroad incumbents
- Hugging Face: Hugging Face is the dominant open-source AI platform for hosting models, datasets, and benchmarks. DMind distributes primarily through Hugging Face; it is both a peer (open LLM publishing, benchmark creation) and a key distribution dependency.
- OpenAI: OpenAI is the leading foundation model lab whose models (GPT-5) DMind itself evaluates. Comparable as the incumbent DMind-3 competes against in the Web3-vertical LLM space, with overwhelming capital, distribution, and capability advantages.
- LangChain: LangChain provides widely adopted developer frameworks for building LLM-powered agents and applications. Comparable to DMind as developer-facing AI tooling infrastructure, though LangChain is general-purpose and not Web3-specialized.
- NEAR Protocol: NEAR Protocol is a Layer-1 blockchain investing heavily in AI integration and on-chain AI agents. Comparable to DMind at the intersection of Web3 infrastructure and AI tooling, though NEAR is a much broader protocol/platform.
Emerging players
- ElizaOS (ai16z): ElizaOS is an open-source framework for building autonomous AI agents in crypto/Web3 environments. Comparable to DMind-3's 'agent ecosystems' use case and Web3-native developer tooling positioning, with similar open-source community-led distribution.
- Alethea AI: Alethea AI builds generative AI and intelligent NFTs on-chain, combining foundation models with Web3-native assets. Comparable to DMind in fusing generative AI with Web3 primitives, though Alethea focuses on character/asset generation rather than financial execution safety.
- Ritual: Ritual is an AI-native crypto network building decentralized inference infrastructure for hosting and serving AI models on-chain. Comparable to DMind in bridging foundation models with Web3 execution environments, though Ritual focuses on infrastructure rather than domain-specialized models.
Direct peers
- ChainGPT: ChainGPT develops AI/LLM tooling specifically for Web3 and crypto use cases (smart-contract generation, audits, trading). Directly comparable to DMind in its AI-for-Web3 positioning, though ChainGPT is more productized and token-incentivized.
- Together AI: Together AI provides open-source foundation model hosting, fine-tuning, and inference APIs. Comparable to DMind as an open-source LLM infrastructure play, though Together AI serves general-purpose workloads rather than Web3-specific verticals.
Regional players
- Render Network: Render Network is a decentralized GPU compute marketplace for AI and rendering workloads. Comparable to DMind in providing Web3-native AI infrastructure, though Render focuses on raw compute rather than domain-specialized LLMs.
Market position
Strengths4 records
Weaknesses4 records
Competitive moat5 records
Key risks6 records
Key highlights6 records
Customer concentration
Dmind social profiles
Digital presenceDmind financial estimates
Financial estimateRevenue estimate
Valuation estimate
Dmind leadership team
Management profileNumber of profiles
Dmind funding detail
Funding detailFunding overview
Funding rounds
Investors
Funding detail is available on the Subscription and Enterprise plan.Contact sales →
Dmind M&A and investment
M&A and investmentM&A
Investments
M&A and investment is available on the Subscription and Enterprise plan.Contact sales →
Frequently asked questions about Dmind
What does Dmind do?
DMind is an open-source AGI research organization that develops and releases large language models (the DMind-1, DMind-2, and DMind-3 series) and AI benchmarks purpose-built for digital finance and Web3. The flagship DMind-3 stack combines an edge intent firewall (Nano), an on-device reasoning engine (Mini), and a policy-governed cloud tier to secure irreversible financial execution against adversarial risks. All products are released under permissive open-source licenses, with API access routed through partners such as NanoGPT and OpenRouter.
Is Dmind a public or private company?
Dmind is a private company. It is classified as unknown and is currently operating.
When was Dmind founded?
Dmind was founded in 2025. It employs 11 to 50 people.
How does Dmind make money?
One revenue line is on record: open-source products and research.
Who are Dmind's main competitors?
Broad incumbents on record are Hugging Face, OpenAI, LangChain and NEAR Protocol. Emerging players are ElizaOS (ai16z), Alethea AI and Ritual. Direct peers are ChainGPT and Together AI. Render Network is listed as a regional player.
Does Dmind have an API?
Yes. DMind offers API access through two providers: NanoGPT, which provides anonymous, private access to ChatGPT and other text and image models with pay-per-prompt pricing using crypto or credit cards; and OpenRouter, which provides a unified API for access to hundreds of AI models through a single endpoint with automatic fallbacks and cost optimization.
What industry is Dmind in?
Dmind's product category is AI Foundation Models for Web3 and Digital Finance. Its primary akta.pro industry code is HDAAACAN, On-Device/Edge Foundation Models (Mobile/Embedded LLMs), with a secondary code of FSAPABAB, Smart Contract Frameworks & SDKs (developer kits, templates, standards). Its NAICS code is 5132 and its SIC code is 7372.