The Thriving Ecosystem of Open Models

2026-06-02 · Source: Tomasz Tunguz · Field: Technology & Digital — Artificial Intelligence & Machine Learning, Emerging Technologies & Innovation · Depth: Intermediate, quick

Summary

Open models have experienced significant growth on OpenRouter since 2025, now comprising 69.1% of the named open-versus-closed token volume, compared to 30.9% for closed models. OpenRouter provides a view into the API frontier where developers can quickly switch and compare models. This surge in usage is driven by new model launches, which attract developer attention and large-scale testing, leading to sustained plateaus in token volume. The open-model ecosystem exhibits rapid innovation and frequent leaderboard changes, similar to closed models. Early leaders like DeepSeek were succeeded by MiniMax and Kimi in late 2025 and early 2026, with subsequent reshuffles by MiMo, Qwen, Alibaba's open-weight family, Hy3, Tencent, and Arcee. This increasing usage and experimentation suggest growing developer willingness to route production traffic to open models.

Key takeaway

AI/ML Engineers choosing production models should recognize the open-model ecosystem's rapid growth and intense competition on platforms like OpenRouter. You should actively test new open-weight releases, such as those from Alibaba or Tencent, to identify cost-effective and performant options. Diversifying your model routing strategy to include these competitive open models can optimize price-performance for your applications.

Key insights

The open-model ecosystem is rapidly expanding, driven by competition and new releases, capturing significant developer token volume on platforms like OpenRouter.

Principles

• Competition drives rapid innovation.
• New model launches boost adoption.
• Developer choice shifts quickly.

In practice

• Monitor OpenRouter for model trends.
• Test new open models upon release.
• Diversify model routing strategies.

Topics

• Open Models
• OpenRouter
• Model Competition
• AI Ecosystem
• Token Volume
• API Frontier

Best for: NLP Engineer, CTO, VP of Engineering/Data, AI Engineer, Machine Learning Engineer, Director of AI/ML

Related on AIssential

• AI data centers — in space As resistance to massive data centers grows on Earth, companies like SpaceX and Google are exploring AI infrastructure in orbit instead — Orbital AI data centers could circumvent terrestrial resource and resistance issues.
• imagine if we had LLMs in the 80s... — The content poses a hypothetical question about the impact of large language models existing in the 1980s.
• Post-Deterministic Distributed Systems: A New Foundation for Trustworthy Autonomous Infrastructure — Autonomous agents challenge deterministic distributed systems, requiring a new model for coordinating diverse, semantically equivalent outcomes.
• Data Management, Analytics and Interoperability — Blockchain-IoT integration enhances data management, analytics, and interoperability through a unified, secure, and real-time architecture.
• AXIOM: A Trust-First Neuro-Symbolic Execution Architecture for Verifiable Mathematical Reasoning — AXIOM uses an LM to canonicalize math problems for a CAS, ensuring verifiable, high-trust reasoning with a dynamic improvement cycle.
• AI Used to Be Generative. Now It’s All About Agents — AI's focus has shifted from generative models to autonomous agents.

Open in AIssential →

Editorial summary, takeaway, and curation by AIssential. Original article published by Tomasz Tunguz.