AI Pulse · Under the Radar

OpenCode, an open-source coding assistant backed by Y Combinator, went from zero to 13 million monthly active users and roughly $40 million in annual recurring revenue in a single year. That’s not a typo. According to Y Combinator’s own account, the surge came largely because Anthropic tightened access for third-party tools inside Claude Code, creating friction that sent developers looking for alternatives.

The lesson here isn’t just about one company’s growth. It’s about what happens when a major platform restricts access and a credible, model-agnostic option exists. OpenCode doesn’t lock you into one AI provider. It works across models, so if Claude gets expensive or restricted, you switch. If OpenAI changes terms, you switch. The architecture itself is the moat, not the model underneath.

For anyone running a business that depends on AI tools, this matters in two ways. First, relying on a single vendor’s API or IDE means you’re exposed when they change the rules. Second, open alternatives are now real competitors with real scale, not just hobbyist projects. Developers vote with their time, and 4.6 million of them chose the open option in a single week.

If you’re building internal AI workflows or agent systems, the same logic applies. The teams that win over the next few years will be the ones that can swap models, route tasks, and adjust without ripping out the entire stack. That’s the kind of flexibility we build into systems like the Omni Command Centre, where the orchestration layer sits above the models, not inside them. When the next clampdown or price hike comes, you want options, not dependencies.