SingularityNET is a decentralized AI model marketplace founded in 2017 by Ben Goertzel, a well-known figure in AI research. Visit the Binance Official Site to check ASI (formerly AGIX) spot prices, download the Official Binance App, and iPhone users should first check the iOS install guide. The short answer: SingularityNET's academic pedigree and Ben Goertzel's AGI vision have attracted a loyal base of long-term believers, but commercial progress has been slow. After AGIX merged into the ASI Alliance in July 2024, the standalone SingularityNET project was effectively reduced to one component within the ASI Alliance. As of April 2026, the merged ASI token (which now includes former AGIX value) carries a market cap of roughly $1.8 billion.
Ben Goertzel is the driving force behind SingularityNET. With more than 40 years of experience in AI research, he was one of the earliest people to popularize the term "artificial general intelligence" (AGI), and he is often mentioned alongside AI pioneers like Marvin Minsky and Ray Kurzweil.
Goertzel's core research focus is OpenCog Hyperon, an AGI architecture that blends symbolic AI (rule-based logic) with neural networks (deep learning). This approach differs from the "pure neural network" path favored by today's mainstream labs such as OpenAI and Anthropic. Goertzel has long argued that neural networks alone cannot achieve AGI, and that symbolic reasoning must be part of the equation.
SingularityNET's design philosophy is essentially about commercializing this research — offering OpenCog's AGI modules as services on the SingularityNET marketplace.
SingularityNET is an AI services marketplace built on the Ethereum mainnet.
Step one: AI model developers publish their models as "Services" on SingularityNET. Each Service comes with an API endpoint, pricing (per call multiplied by a unit price), and usage examples.
Step two: users browse the marketplace through the SingularityNET DApp, pick the AI service they need, and pay for each call using AGIX/ASI tokens.
Step three: the AI model runs on the developer's own servers and returns the result. Token settlement happens automatically through SingularityNET's on-chain contracts.
The whole flow works a bit like an "AI service marketplace" — any developer can list a model, and any user can pay to call it.
SingularityNET's real-world platform numbers to date are worth a closer look.
| Metric | Data |
|---|---|
| AI services listed | 70+ |
| Monthly active calls | ~20,000 (2025 data) |
| Main service types | Text, image, voice, AGI experiments |
| Main customers | Academic researchers, independent developers |
| Monthly platform revenue (token settlement) | Roughly $50,000-$100,000 equivalent |
These numbers show that SingularityNET is still far from being a mainstream AI services platform. Twenty thousand monthly calls is negligible compared to OpenAI's API, which handles billions of calls per day. In practice, SingularityNET occupies the niche of a "boutique AI experimentation marketplace," serving mostly academic researchers and AGI enthusiasts.
AGIX was merged into the ASI Alliance in July 2024. Under the merger plan, AGIX holders exchanged their tokens for FET (later renamed ASI) at a fixed ratio of roughly 0.4225.
After the merger, the SingularityNET project was effectively downgraded to a component of the ASI Alliance. Ben Goertzel serves as CEO of the ASI Alliance, but the SingularityNET platform's product independence has been diluted — most new AI features are now released under the unified ASI Alliance banner (such as the ASI:One LLM).
| Dimension | AGIX Before Merger | ASI After Merger |
|---|---|---|
| Token market cap | $800 million | $1.8 billion (merged token) |
| Total supply | 1.2 billion AGIX | 2.75 billion ASI |
| Project independence | Fully independent | Alliance member |
| Ben Goertzel's role | SingularityNET CEO | ASI Alliance CEO |
OpenCog Hyperon is the centerpiece of SingularityNET's academic pedigree. It's an AGI framework developed by Ben Goertzel's team, aiming to "achieve genuine general intelligence through a hybrid of symbolic reasoning and neural networks."
| OpenCog Hyperon Feature | Details |
|---|---|
| Development language | C++ plus Atomese (a proprietary symbolic language) |
| Core components | AtomSpace (knowledge graph), PLN (probabilistic logic) |
| Current progress | Version 0.6 (alpha) |
| Release timeline | Version 1.0 planned for 2026-2027 |
OpenCog Hyperon's actual product is still orders of magnitude behind GPT-4 or Claude 3. Goertzel's team brings 40+ years of accumulated research and a distinctive methodology for symbolic reasoning, but whether that translates into a commercially viable product remains an open question.
Controversy one: weak long-term delivery. SingularityNET raised $36 million in its 2017 ICO and promised a usable AI marketplace within 2-3 years. Eight years later, the platform still handles only about 20,000 monthly calls — badly out of step with its original roadmap.
Controversy two: out of step with the AI mainstream. Ben Goertzel has long argued that "pure neural networks cannot achieve AGI," yet OpenAI's GPT series has already displayed intelligence far beyond what was expected. SingularityNET's technical direction is not a consensus view within mainstream AI academia.
Controversy three: loss of independence after the merger. After AGIX merged into ASI, SingularityNET's brand and product independence dropped significantly. The "AGI concept" value that original AGIX holders bought into has been diluted into the broader "AI Agent economy" narrative.
Opportunity one: an AGI breakthrough. If OpenCog Hyperon achieves some kind of AGI-level breakthrough between 2026 and 2028, SingularityNET — as a key component of the ASI Alliance — could see returns far exceeding its current valuation. This is Ben Goertzel's core bet.
Opportunity two: the long-tail AI market. SingularityNET has a first-mover advantage in "niche, specialized, academic AI services." If centralized AI providers like OpenAI and Anthropic focus mainly on general-purpose large models, there may still be room for long-tail specialized AI models, and SingularityNET could become the infrastructure for that segment.
Opportunity three: decentralization of AI. If regulation or geopolitics pushes toward decentralizing AI (to avoid a single company controlling AGI), SingularityNET's decentralized design carries a structural advantage.
SingularityNET and Bittensor are both decentralized AI projects, but they take completely different paths.
| Dimension | SingularityNET | Bittensor |
|---|---|---|
| Model | AI service marketplace | AI model mining (PoW-like) |
| Incentive target | Developers listing services | Miners submitting models |
| Chain | Ethereum mainnet | Proprietary Substrate chain |
| Tokenomics | 2.75 billion (low inflation post-merger) | 21 million hard cap plus halvings |
| Actual on-chain activity | ~20,000 calls/month | 12,000+ miners |
Bittensor clearly leads SingularityNET on objective metrics. Bittensor's "halving plus subnet" model directly drives miner competition, generating substantial real on-chain activity; SingularityNET's "marketplace model" depends on external demand and lacks a built-in incentive engine.
The long-term believer's view: buy ASI (which now includes SingularityNET's value) as a long-term bet on Ben Goertzel's team and an eventual AGI breakthrough. This is a 5-10 year time horizon, with expected returns tied directly to the probability of AGI actually happening.
The short-term trader's view: ASI correlates closely with the AI sector, so it can be allocated during AI-token sector rotations. Short-term volatility is severe and strict stop-losses are essential.
The reason to avoid it: looking purely at fundamentals — actual on-chain activity and commercial progress — SingularityNET's financial performance doesn't support its current valuation.
Q: Can AGIX still be traded?
No. After AGIX merged into ASI, the original AGIX token contract stopped circulating. Exchanges like Binance and OKX completed the conversion automatically long ago. If you still hold AGIX in a self-custody wallet, you'll need to manually convert it through the ASI Alliance's official website.
Q: What does Ben Goertzel do at the ASI Alliance?
He serves as CEO of the ASI Alliance, overseeing the development direction of the ASI:One LLM, the AGI research roadmap (OpenCog Hyperon), and external academic partnerships.
Q: Is the SingularityNET platform still running?
Yes, but activity is low. The platform URL (singularitynet.io) is still accessible, and the 70+ AI services remain online. But monthly active calls sit around 20,000, making it a niche market.
Q: Can AGI actually be achieved?
There's no academic consensus. The mainstream view held by OpenAI and Anthropic is that scaling up neural networks plus reinforcement learning will lead to AGI, expected within 5-15 years. Ben Goertzel's minority view holds that symbolic AI must be combined in for true AGI to emerge. Neither path has clear proof of success yet.
Q: What's the relationship between ASI:One and OpenCog Hyperon?
ASI:One is the ASI Alliance's practical large language model (Transformer-based), used for Agent tasks. OpenCog Hyperon is Ben Goertzel's long-term AGI research project (based on a hybrid of symbolic reasoning and neural networks). The two use different technical approaches and there's currently no product-level integration between them.
Q: How good are the AI services on SingularityNET?
Quality varies widely. Most services are small models built by individual developers, far below the quality of the OpenAI API. A handful of services — such as OpenCog's symbolic reasoning tools — have academic value, but limited commercial utility.
Q: What is a long-term ASI holder actually betting on?
They're betting on any one of three things happening: an AGI breakthrough within 5-10 years (Goertzel's path), the AI Agent economy taking off (Fetch.ai's path), or AI data marketplaces becoming core infrastructure (Ocean Protocol's path). If any one of these three succeeds, ASI has room for a multi-fold price increase.
Q: What happens if Goertzel leaves?
Ben Goertzel is over 60 years old, and he may gradually step back from day-to-day operations in the coming years. If he leaves his role as CEO of the ASI Alliance, ASI's "AGI academic narrative" would weaken, which could affect the price.
SingularityNET has had limited product delivery over eight years of operation, and its 20,000 monthly calls are badly out of proportion with its token market cap. Ben Goertzel's AGI roadmap is not a consensus view within mainstream AI academia, and its long-term chance of success is hard to assess. This article is compiled from public data and does not constitute investment advice. Please read our Risk Disclaimer.
Data sources: SingularityNET official blog, ASI Alliance announcements, CoinGecko, Etherscan (April 2026).