CAIA Logo
CAIBA
CAIBA’s Momentum Continues: Welcoming Our Next Wave of Pioneers

CAIBA’s Momentum Continues: Welcoming Our Next Wave of Pioneers

Jul 17, 2025

One Month In: CAIBA Expands, Benchmarks Scale, and the Crypto-AI Frontier Advances

The Crypto AI Benchmark Alliance (CAIBA) was born to accelerate safe, open innovation at the crossroads of blockchain and artificial intelligence. Just four weeks since launch, CAIBA is proving that community-led standards can move fast and fair.

Benchmarks are growing. Partnerships are deepening. And the path toward verifiable, open-source Crypto AI is coming into focus.

📈 First-Month Highlights

Since our launch on June 3, 2025, we’ve seen momentum build across every front:

  • Three open benchmarks released:
    • Crypto Deep Research (formerly CAIA, or Crypto AI Agent)
    • OCE (Onchain Execution)
    • Crypto NER (Named Entity Recognition)
  • Four analytic reports published, covering agent and model performance across all three benchmark tracks.
  • Transparent evaluation framework launched: all tasks, results, and criteria are open-sourced for anyone to inspect, replicate, and contribute to.

Together, these early benchmarks are setting a new precedent: open, verifiable benchmark standards for the Crypto AI era.

🤝 Meet Our Second Cohort of CAIBA Partners

The interest in CAIBA has been overwhelming and inspiring. Today, we’re excited to welcome eight exceptional new partners to the Alliance:

  • Kite AI — Architecting a purpose-built L1 for the agentic internet with native support for real time payments, programmable governance, and cryptographic trust.
  • GM Agents — All your AI agents in one app — use more, earn more.
  • Flock — Pioneering decentralized AI training via federated learning on blockchain rails.
  • Codatta — Building a decentralized Knowledge Layer, tailored for post-training agent fine-tuning through high-quality community data.
  • Buzzing — Real-time betting engine and truth oracle across social channels, turning noise into insight.
  • Nexus — A world supercomputer to enable the AI economy.
  • Ormi Labs — A hyperscale data platform delivering live, historical, and AI-enriched data at sub-second latency.
  • DMind — An open-source AGI institute bridging real world crypto data with foundational AI research and open benchmarks.

Each of these teams brings expertise that sharpens CAIBA’s foundation—from benchmark design to production grade evaluation and knowledge curation. Expect new research articles, joint code contributions, and community collaborations in the weeks ahead.

🛠️ What’s Next: Benchmark Evolvement, Protocol Expansion & Scoreboard Automation

Here’s what’s coming soon:

  • Crypto Deep Research Benchmark v1.0 — Introducing more agent testing with task decomposition and a live scoreboard for real time evaluation.
  • OCE Expansion — Benchmarks to extend to more networks and protocols, deepening execution surface coverage.
  • Research & Dev Collaborations — New replication studies, reproducible agent tests, and agent tuning challenges.

If you’re building Crypto AI agents, training knowledge systems, or investing in Crypto AI infrastructure, this is your signal to get involved.

👇 How to Plug In

We’re building open rails for agentic intelligence in crypto. Here’s how to join the movement:

  • Read the benchmark specs and reports → caiba.ai
  • Apply to join CAIBA → Bring your models, datasets, infra, or ideas.
  • Submit prompts, pull requests, or replications → Every contribution strengthens the benchmark layer.

Together, we’re laying down transparent, auditable, community governed benchmarks that will power the next decade of Crypto AI.