compinio.io

AI-powered crypto research

AI analysis for crypto microcaps.
Deep. Verifiable.

Every published token has been through an algorithmic screen, an AI deep-dive, and a Twitter quality check. The analysis is frozen at publish time and never rewritten.

Latest picks ↓How it works

Latest picks

2 / 2
SERV
SERVOpenServ
Published May 14, 2026
3/5Risky

OpenServ is a platform where non-technical users can assemble teams of AI agents to handle business tasks and workflows together without writing code.

+45.0%since published+103.7%vs sector · 180d$0.0481
Read full analysis ↓Updated 2m ago

OpenServ is outperforming its sector massively (up 138% over 7 days while AI-agent peers crawl) suggesting renewed interest in the AI-agent trend; the leading comparable, Virtuals Protocol, is currently worth around $400M while OpenServ sits near $25M, leaving room to grow if agent-orchestration platforms regain traction, though as a small project it can also lose most or all of its value.

Verdict — Real AI-agent platform riding a strong momentum reversal, but value capture and competitive position remain unproven.

Hypothetical scenariosarithmetic, not a forecast

If it reached the valuation of comparable projects: 20×, $100 hypothetically becomes $1.9K.

  • vs Virtuals Protocol $912.01M FDV20×

Downside is total. Microcap tokens go to zero far more often than they multiply. These figures are arithmetic comparisons of current valuations — not predictions, not targets, not advice. You could lose everything.

Price
$0.0481
Market value
$36.46M
Full valuation
$48.10M
24h volume
$3.35M
Artificial Intelligence (AI)Ethereum EcosystemBase EcosystemAI AgentsAI Frameworkx402 Ecosystem
𝕏 4/5Published May 14, 2026@openservai ↗Share image ↗

Real AI-agent platform riding a strong momentum reversal, but value capture and competitive position remain unproven.

Real project with a coherent product, strong technical momentum bucking a weak sector, and reasonable market cap — but novelty is modest, competition is severe, and 23% supply overhang plus unverified usage cap the conviction. Push to tier 4 with evidence of actual platform usage and token sink mechanics; drop to tier 2 if the rally fades and reveals it was just sector beta on thin liquidity.

Novelty

3/5

Multi-agent orchestration with cross-framework interoperability is a real product angle, but the space (Virtuals, ai16z/ElizaOS, Olas, Fetch) is crowded. The no-code team assembly is a reasonable differentiator but not a defensible moat.

Technicals

4/5

Massive outperformance versus sector: +138% 7d and +138% 90d while AI-sector medians are flat to negative. Volume of $8M against $25M mcap is healthy. Still -50% on the year, so this is a recovery off lows rather than a fresh ATH breakout — momentum is real but extended short-term.

Whitepaper

Whitepaper URL is a Docsend link, content not readable in this analysis; score withheld. Note that hosting docs behind Docsend rather than a public PDF is a mild negative signal for transparency.

Competitors

  • Virtuals Protocol, Larger ecosystem and agent launchpad traction · $912.01M FDV
  • ai16z / ElizaOS, Strong open-source dev mindshare and framework adoption
  • Olas Network, Production multi-agent infra with real on-chain usage · $17.72M FDV
  • Fetch.ai, Mature agent network, exchange listings, enterprise relationships

Risks

  • 23% of supply still to unlock — dilution pressure ahead
  • Crowded AI-agent sector with better-funded incumbents
  • Unclear token value capture; SERV utility may be weak
  • Short-term price extended after +138% week, mean-reversion likely
  • Product traction not verifiable from data provided
AI-generated analysis
MOMUS
MOMUSMomus ai
Published May 13, 2026
3/5Risky

MOMUS is an automated program that places bets on Polymarket — a website where people wager on real-world events like elections — and posts its trades and reasoning publicly on X.

+48.3%since published+43.1%vs sector · 7d$0.000123
Read full analysis ↓Updated 2m ago

Worth roughly $61,000 today, MOMUS could climb toward the value of similar AI trading bot tokens like Aixbt, worth around $80 million, if its bot proves to genuinely make money over time, though at this size the token might also lose all its value.

Verdict — Speculative microcap pick — 5-day-old Solana AI agent claiming autonomous Polymarket trading; plausible narrative but unverifiable, no GitHub, tiny float.

Hypothetical scenariosarithmetic, not a forecast

If it reached the valuation of comparable projects: 7,400×, $100 hypothetically becomes $739K.

  • vs Virtuals Protocol $912.01M FDV7,400×

Downside is total. Microcap tokens go to zero far more often than they multiply. These figures are arithmetic comparisons of current valuations — not predictions, not targets, not advice. You could lose everything.

Price
$0.000123
Market value
$98.7K
Full valuation
$123.4K
24h volume
$39.7K
Artificial Intelligence (AI)Solana EcosystemAI Agents
𝕏 4/5Published May 13, 2026@momus_ai ↗Share image ↗

Speculative microcap pick — 5-day-old Solana AI agent claiming autonomous Polymarket trading; plausible narrative but unverifiable, no GitHub, tiny float.

Links resolve, name/symbol/domain are consistent, category is specific, no scam-pattern floor triggers — that clears tier 2. But no GitHub, no training-data familiarity, derivative thesis, and unverifiable agent claims keep it from tier 4. Would move to 4 with a verifiable on-chain track record of the agent's Polymarket positions plus a real public repo; drops to 2 if the agent's 'public reasoning' on X turns out to be sparse or fabricated.

Novelty

2/5

Autonomous AI agents trading prediction markets is a crowded 2024-25 narrative (Aiko, several Polymarket bot agents already public on X). Publishing reasoning + losses transparently is a nice touch but not a moat. Nothing here is technically distinctive.

Technicals

3/5

Only 5 days of price data; 7d/30d/90d windows literally unassessable — neutral by default. +89% since launch and $38K daily volume on $61K mcap is high turnover (~60%), which at this age is normal post-launch churn but watch for distribution. Sector median 7d is +6%, so AI narrative is mildly supportive.

Whitepaper

Whitepaper URL (GitBook) provided but content not readable in this analysis; score withheld. GitBook docs at microcap stage are usually thin marketing pages — verify before weighting.

Competitors

  • Aixbt, Established alpha-agent brand with real social traction and integrations
  • Polymarket, The underlying venue; any edge Momus claims is captured by Polymarket-native traders first
  • Virtuals Protocol, Full agent-launch platform with broader distribution and tokenomics framework · $912.01M FDV

Risks

  • No GitHub — claims of autonomous agent unverifiable; could be a human trader with a Twitter bot
  • 5 days old, tiny mcap — high rug/abandonment probability
  • Deflationary 'every task burns tokens' tokenomics rarely produce meaningful burn at low usage
  • Heavy competition from established AI-agent tokens with bigger distribution
  • 20% of supply not yet circulating — unlock/dump risk
AI-generated analysis
Not financial advice. Informational only. Microcap crypto is extremely high risk, most tokens lose substantial value. Do your own research.

Method

How a token becomes a pick

Every published pick has been through four sequential filters. None of this is an opinion call until the very last step, and even that step is bounded by a published rubric.

  1. 1

    Algorithmic screen

    CoinGecko universe filtered by hard rules: market cap band (microcap $1–$100K, screen $100K–$1B), minimum 24h volume, maximum age, and exclusions for memes, stablecoins, gaming infra, parasitic L2s.

  2. 2

    Volume Quality (VQ)

    A 1–5 composite score for how much of the reported volume is real: green-tier exchange share, listing diversity, weighted spread, and roundness anomalies. Sub-2 flags likely wash; the pick still publishes, with the flag in plain view.

  3. 3

    AI deep-dive

    Survivors are sent to an AI model with their technicals, sector context, and project description. Output: novelty, technicals, whitepaper, competitors (with current FDVs from our DB), risks, and a 1–5 tier. A scam-pattern floor caps the tier at 2 when name / links / claims don't line up.

  4. 4

    Twitter verification

    Before going public, a 1 to 5 score for the project's Twitter: content quality, engagement organicness, founder voice. A draft without this score never reaches this page.

What we can't verify

  • Who the team really is — we read whitepapers and socials, we don't run KYC.
  • On-chain contract authorship and rug risk (coming later — Etherscan + Helius).
  • Whether a token will go up. Receipts audit our decision quality, not returns.