University of Kwazulu-Natal

Llm Rankings Openrouter

Bonisiwe Shabane
-
llm rankings openrouter

43 openrouter models are actively benchmarked with 14641 total measurements across 14316 benchmark runs. llama-3.2-1b leads the fleet with 113.00 tokens/second, while gpt-4.1-nano delivers 65.60 tok/s. Performance varies by 72.3% across the openrouter model lineup, indicating diverse optimization strategies for different use cases. Avg time to first token across the fleet is 829.07 ms, showing good responsiveness for interactive applications. The openrouter model fleet shows varied performance characteristics (44.4% variation coefficient), reflecting diverse model architectures. OpenRouter’s comprehensive rankings let you compare language models across 12 categories, including Programming, Translation, Marketing, Roleplay, and more.

OpenRouter provides insights into how models perform over different periods. Here’s a breakdown for the Programming category: While Gemini 2.5 Pro is gaining ground, Claude 3.7 Sonnet remains the top performer of the day. This consistency highlights Claude's reliability for programming tasks. GPT-4o-mini took the weekly crown, even as Quasar Alpha (OpenAI’s Q* model) briefly appeared as a free option before disappearing. This shows how quickly the landscape can shift.

Note that Quasar Alpha was offered as a free model for a while hence the heavy usage. This report summarizes the usage and market share trends for Large Language Models (LLMs) and their corresponding providers and applications, based on data from the OpenRouter Rankings spanning October 2024 to October 2025. The total usage across all models has shown a steep increase over the observed period, with cumulative tokens surpassing T tokens by September 15, 2025. The Grok and Claude models demonstrate significant dominance in overall token usage: The provider market share data, spanning October 13, 2024, to September 14, 2025, shows x-ai as the clear leader: The recent data (September 16 - October 10) and specific category rankings highlight the model's specialized performance.

OpenRouter’s LLM Rankings page aggregates real-time production data from its unified API to show which models, labs and public applications consume the most tokens. Views include Top today/this week/this month, market-share pie charts, and per-use-case model categories. Developers and researchers can quickly identify trending models, price-to-performance leaders and community adoption patterns without running their own telemetry. Track, rank and evaluate open LLMs and chatbots Explore and compare advanced language models on a new leaderboard Track, rank and evaluate open LLMs and chatbots

Explore and compare advanced language models on a new leaderboard Track, rank and evaluate open LLMs and chatbots Ryan MacLean provides a comprehensive exploration of OpenRouter, demonstrating how it serves as a unified interface for accessing multiple LLM providers and solving enterprise challenges around model selection and routing. The episode covers OpenRouter's real-time model rankings, showing Claude Sonnet 4 leading usage statistics, and walks through practical implementation using the familiar OpenAI SDK. Ryan demonstrates the platform's Request Builder for interactive testing, explains the various pricing tiers including free options with data training caveats, and details privacy settings ranging from endpoints that may train on data to... The discussion highlights how OpenRouter addresses vendor lock-in issues by providing access to models from different cloud providers (AWS, Azure, Google Cloud) through a single API, making it particularly valuable for enterprises building chat...

Unified interface for accessing multiple LLM providers with routing and orchestration capabilities Complete API documentation with code samples for Python, JavaScript, and shell Real-time rankings and usage statistics for available models Official OpenAI Python client library, compatible with OpenRouter API The leaderboard shows a mix of models from different providers at the top. Grok Code Fast 1, Gemini 2.5 Flash, and Claude Sonnet 4.5 are the top three models.

There is a strong presence of Google's Gemini models in the top 10, along with models from Anthropic (Claude), DeepSeek, and Grok. The usage numbers are in trillions of tokens, indicating substantial usage across the platform. This ranking is automatically updated weekly via GitHub Actions using screenshot analysis and AI. Data source: OpenRouter Rankings Analysis powered by Google Gemini 2.5 Pro

People Also Search

43 Openrouter Models Are Actively Benchmarked With 14641 Total Measurements

43 openrouter models are actively benchmarked with 14641 total measurements across 14316 benchmark runs. llama-3.2-1b leads the fleet with 113.00 tokens/second, while gpt-4.1-nano delivers 65.60 tok/s. Performance varies by 72.3% across the openrouter model lineup, indicating diverse optimization strategies for different use cases. Avg time to first token across the fleet is 829.07 ms, showing goo...

OpenRouter Provides Insights Into How Models Perform Over Different Periods.

OpenRouter provides insights into how models perform over different periods. Here’s a breakdown for the Programming category: While Gemini 2.5 Pro is gaining ground, Claude 3.7 Sonnet remains the top performer of the day. This consistency highlights Claude's reliability for programming tasks. GPT-4o-mini took the weekly crown, even as Quasar Alpha (OpenAI’s Q* model) briefly appeared as a free opt...

Note That Quasar Alpha Was Offered As A Free Model

Note that Quasar Alpha was offered as a free model for a while hence the heavy usage. This report summarizes the usage and market share trends for Large Language Models (LLMs) and their corresponding providers and applications, based on data from the OpenRouter Rankings spanning October 2024 to October 2025. The total usage across all models has shown a steep increase over the observed period, wit...

OpenRouter’s LLM Rankings Page Aggregates Real-time Production Data From Its

OpenRouter’s LLM Rankings page aggregates real-time production data from its unified API to show which models, labs and public applications consume the most tokens. Views include Top today/this week/this month, market-share pie charts, and per-use-case model categories. Developers and researchers can quickly identify trending models, price-to-performance leaders and community adoption patterns wit...

Explore And Compare Advanced Language Models On A New Leaderboard

Explore and compare advanced language models on a new leaderboard Track, rank and evaluate open LLMs and chatbots Ryan MacLean provides a comprehensive exploration of OpenRouter, demonstrating how it serves as a unified interface for accessing multiple LLM providers and solving enterprise challenges around model selection and routing. The episode covers OpenRouter's real-time model rankings, showi...