Zeno Divergent is a research-stage ‘model of models’ that evaluates forecasts across macroeconomic, weather, market, and climate domains. Its official website (zenodivergent.dev) describes a system that collects existing forecasts, measures disagreement (‘surprise’), and flags when collective judgment becomes fragile — with all forecasts locked in advance and scored publicly against outcomes. The project operates on Base (contract 0xE883eFe16E97aFD09550bDB6BC977baD14fDf3Cd) and Solana (token DVska6Y1PohdNJD2Gt6DwQD1EypfEvcmrFfdttCiRkbf), with explorers including Solscan and BaseScan. A whitepaper is hosted at the same domain, and GitHub activity is attributed to mbarbosa30. According to API-sourced metadata, ZENO has a max supply of 1 billion tokens and zero circulating supply as of 2026-09-25; no treasury assets, liabilities, burn schedule, or runway are disclosed. Key evaluation questions remain: What capital supports ongoing development? Is there verifiable evidence of deployed forecasting infrastructure beyond v0.29 previews? How are token economics tied to utility, given no live challenge programme or holder rights are described?
Zeno Divergent positions itself as a research-stage ‘model of models’ that evaluates forecast reliability across domains like macroeconomics, weather, and markets. Its core product—publicly scored, timestamped forecasts—is already operational: the website states 3,895 forecasts are locked in advance and 2,305 scored against reality, with live daily scoring since at least 2026-09-26 (zenodivergent.dev). The whitepaper and documentation are referenced but not supplied; no audit, third-party validation, or performance benchmarking beyond internal baselines (e.g., GBM v0.15.0) appears in evidence.
The $ZENO token has no current functional use. Per CoinGecko, it is ‘intended to support open, verifiable forecasting challenges’—but the challenge programme is explicitly ‘not live yet’. No token-dependent user action (e.g., staking to submit forecasts, governance over Arena rules, fee capture from scoring services) is described or implemented. Circulating supply is zero, and no value flow—fee, revenue share, redemption, or enforceable rights—is traceable in the evidence. The token’s economic mechanism remains hypothetical: future utility depends entirely on unlaunched infrastructure. This absence of token necessity or verified value capture triggers the rubric’s hard cap: without verifiable supporting evidence for the primary thesis (i.e., that $ZENO enables or sustains the system), confidence is low and scores above 6 are disallowed.
No material contradictions or critical asset exposures are verified in the snapshot.
Overall score: 5/10 Confidence: Low
Zeno Divergent presents itself as a research-stage ‘model of models’ that evaluates forecast reliability across macro, weather, markets, and hazards—scoring forecasts publicly and locking them in advance. Its website documents 3,895 locked forecasts and 2,305 scored against reality, with live daily scoring and domain expansion from v0.28 to v0.29 (adding verified weather data). The project operates on Base and Solana, with contracts published and explorers linked, and maintains GitHub, Twitter, and Telegram channels. However, the primary thesis—that it reliably detects forecast fragility to inform decisions—lacks verifiable evidence in the snapshot: no performance benchmarks against baselines are shown beyond claims like ‘Beat GBM v0.15.0’, no third-party validation or audit is cited, and the whitepaper URL returns no accessible content. Token supply data shows zero circulating supply and no treasury or incentive mechanism disclosed, making acquisition, retention, and organic participation impossible to assess. Community activity is asserted but not evidenced: no engagement metrics, contributor history, or moderation records appear. The API source explicitly labels the project as ‘research-stage’ and notes the challenge program ‘is not live yet’. With no independent verification of model behavior, user traction, or operational continuity—and all evidence originating from self-reported sources—the assessment is constrained by narrow, uncorroborated, and non-current proof.
Overall score: 6/10 Confidence: Low