2 min read
|
Saved February 14, 2026
|
Copied!
Do you care about this?
The article examines the competitive landscape of open-source and proprietary AI models, highlighting that proprietary providers maintain pricing power despite cheaper alternatives. Open-source models have stabilized at about 22-25% market share, while programming use cases dominate among leading providers. Retention rates vary significantly, with some models showing stronger user engagement than others.
If you do, here's more
OpenRouter's data reveals a clear divide in the AI market, where open-source models, despite being significantly cheaper, have maintained only 22-25% market share over the past year. A brief spike to 35% occurred in mid-2025 due to the rise of Chinese models, but demand for proprietary providers remains strong. These companies retain pricing power, especially for mission-critical applications, while open-source models attract cost-sensitive users.
The landscape for open-source models is shifting as well. DeepSeek's share dropped from nearly 80% to 40% in early 2025, as competitors like Qwen and other Chinese models gained traction. Programming has emerged as a key use case, with Anthropic and xAI reporting 60% and 45% of their usage in this area, respectively. Roleplay is also growing rapidly, with DeepSeek capturing 80% of its volume in this segment, driven by consumer price sensitivity.
Retention rates reveal a stark reality; most models experience a churn of 60-70% within the first month. However, those that find product-market fit, like Claude 4 Sonnet and Gemini 2.5 Flash, achieve better retention rates of 40-50% in the first month. In contrast, models like GPT-4o Mini and DeepSeek R1 struggle, retaining only 25-35% of users. Ultimately, enterprises are willing to pay for precision, while consumers remain hesitant to spend on entertainment, leaving proprietary providers unchallenged on price in their segment.
Questions about this article
No questions yet.