OpenRouter
@OpenRouter
New public rankings: Audio Input! @GoogleDeepMind's Gemini models take the top 7 (!!) slots this week, with Gemini 3 and 2.5 Flash models processing >50% of prompts. https://t.co/FTOTOk61jP
5retweets69likes
View on XOpenRouter Audio Rankings Reveal Google Gemini Dominance in Multimodal Usage
OpenRouter· Updated
OpenRouter's new Audio Input leaderboard shows Google's Gemini models capturing the top seven spots and over 50% of total audio prompts. While competitors like OpenAI and Mistral appear in the top ten, developers are overwhelmingly choosing Gemini's Flash variants for production audio workloads.
OpenRouter, a unified API for accessing hundreds of AI models, launched a new public leaderboard category for audio input usage. The data reveals that Google currently dominates the audio space, with its Gemini models occupying the top seven positions. The
Gemini 3 Flash Preview leads with a 43.4% share.- Gemini 3 Flash Preview share
- 43.4%
- Gemini 2.5 Flash share
- 16.6%
- Gemini 3.1 Pro Preview share
- 13.7%
- Gemini top slots
- 7 of 10
- Combined Flash prompt share
- >50%
- GPT Audio Mini share
- 2.2%
This shift highlights a clear market preference for native audio reasoning in production. While we previously saw Gemini 3.1 Flash Live top performance benchmarks, this usage data confirms that developers are prioritizing Google's low-latency Flash architecture over alternatives from OpenAI or Mistral.
You can now use these rankings to benchmark model selection for voice-enabled agents. The leaderboard is updated weekly based on real-world token volume, providing a transparent view of which models are deployed at scale. Access to all ranked models is available through the OpenRouter API.
Still wondering? A few quick answers below.
OpenRouter Audio Input rankings are a public leaderboard that tracks the total number of audio prompts processed by different AI models on the platform. The data is based on real usage from millions of users and developers, providing a transparent view of which models are actually being used for audio-to-text and audio reasoning tasks.
Google Gemini models currently dominate the audio input category, taking the top seven spots on the leaderboard. Gemini 3 Flash Preview is the most used model, accounting for 43.4% of all audio prompts. Other popular choices include Gemini 2.5 Flash and Gemini 3.1 Pro Preview, which together process more than half of the platform's audio traffic.
While Google Gemini models hold the top seven positions, other providers have a smaller presence in the top ten. Mistral's Voxtral Small ranks eighth with 2.5% of prompts, and OpenAI's GPT Audio Mini ranks ninth with 2.2%. The data shows that the majority of audio prompts on the platform are currently being routed to Google's multimodal models.
The rankings on OpenRouter are updated weekly to reflect the latest usage trends and model performance. These leaderboards track various metrics, including total token volume, market share, and specific input types like audio or images, allowing developers to see which models are gaining traction in the rapidly evolving AI landscape.
Yes, all models featured in the rankings, including the top-performing Google Gemini series, are accessible through OpenRouter's unified API. This allows developers to integrate multiple models from different providers into their applications using a single integration, making it easier to switch between models or route requests based on current performance and usage data.
Every HeadsUpAI update is written based on its original source and reviewed before it's published. Read our editorial standards →
