Expanding Our Intelligence: WildChat-4.8M Joins LLM Radar

We’re thrilled to announce a major expansion to LLM Radar: the integration of Allen Institute for AI’s massive WildChat-4.8M dataset!

By combining WildChat’s rich, multi-turn interaction logs with our existing daily token data from OpenRouter, LLM Radar now offers an unprecedented level of real-world intelligence into how large language models are actually being used.

The Strategic Advantage: Beyond Basic Token Counts

Until now, tracking LLM momentum mostly meant looking at top-level API token volume. While high-level usage metrics are critical, they only tell part of the story.

By layering WildChat-4.8M alongside our OpenRouter dataset, LLM Radar bridges the gap between raw API consumption and real-world conversation dynamics:

• Real-World Chat Dynamics: Gain visibility into actual user prompts, turn frequencies, language distributions, and multi-turn conversation lengths.
• Deeper Historical Context: Trace model usage patterns over longer timeframes across key architectural shifts and model releases.
• Model-Specific Granularity: Inspect detailed telemetry for specific model families—such as GPT-4 variants, open-weight models, and reasoning architectures—to see how user interaction styles differ across models.

Why This Matters for Developers & Researchers

Evaluating AI models isn’t just about benchmark scores or published API prices—it’s about how models perform in production under real-world prompt distributions.

With this new dataset integration, LLM Radar users can:

1. Refine Cost & Token Estimates: Understand actual input-to-output token ratios across real user chats to make the LLM Token Calculator even more accurate.
2. Analyze User Behavior: See how prompt lengths, system instructions, and multi-turn sessions scale in practice.
3. Make Data-Driven Architecture Decisions: Choose models based on proven real-world adoption, performance efficiency, and cost viability.

What’s Next?

The addition of WildChat-4.8M marks the next step in our vision to make LLM Radar the ultimate data-driven dashboard for the AI ecosystem. We will continue expanding our datasets, refining our analytics UI, and building tools that make exploring AI models simple, accessible, and actionable.

Explore the new data today on LLM Radar and test your prompts on the LLM Token Calculator!