Overview
LMSYS Chatbot Arena Leaderboard is a crowdsourced evaluation platform that ranks large language models (LLMs) based on real human preferences instead of synthetic benchmarks. Users chat with two anonymous models side by side and vote for the better answer, creating pairwise comparisons that reflect practical, everyday quality. With over one million human votes, the leaderboard applies the Bradley–Terry model and Elo-style ratings to generate a robust, statistically grounded ranking of open-source and proprietary LLMs.
The leaderboard offers a transparent view of how models perform across a wide range of tasks, including coding, reasoning, writing, and general conversation. Researchers, developers, and product teams can use these ratings to select the most suitable model, compare new releases, and track rapid progress in the LLM ecosystem. Because the platform is continuously updated with new battles and models, it captures evolving trends rather than static scores.
LMSYS Chatbot Arena Leaderboard is completely free to use via the web, making state-of-the-art LLM evaluation accessible to individuals, startups, and enterprises alike. Whether you are choosing a model for your next product, benchmarking your own LLM, or simply exploring how top chatbots stack up, the leaderboard provides an unbiased, community-driven reference backed by large-scale human judgments.
LMArena Text Leaderboard
Top 10 LLMs by current Elo / Bradley–Terry scores from LMArena human-preference battles. Click a model name to see its ToolHub family page; ↗ links to the official source.
| # | Model | Elo Score |
|---|---|---|
| 1 | Anthropic | 1509 |
| 2 | Anthropic | 1504 |
| 3 | Anthropic | 1502 |
| 4 | Anthropic | 1499 |
| 5 | Anthropic | 1494 |
| 6 | Meta | 1487 |
| 7 | 1486 | |
| 8 | 1486 | |
| 9 | Anthropic | 1484 |
| 10 | OpenAI | 1481 |
- Elo gap from #1 to #10 is just 28 points.Differences under ~10 points are often within noise — treat the entire top 10 as one tier and pick by task fit and cost, not by who’s “smartest” on this single chart.
- Pick by task, not by overall rank:coding, refactoring, and long-document analysis → Claude; general utility + voice + image → ChatGPT; Workspace integration + multimodal → Gemini; self-hosted or zero-cost → Meta AI / Llama.
- Are “thinking” variants worth it? Thinking models excel at hard reasoning but cost more latency and tokens. Use the regular variant by default; switch to thinking only for tough coding, math, or multi-step problems.
- This is the Text leaderboard only. Coding battles, web dev, vision understanding, and long context each have their own sub-arena — see WebDev / Vision / Coding tabs on the official LMArena site.
Pricing
Free
Detailed plans have not been confirmed in our catalog. Check the official website for current limits and billing terms.
Visit WebsitePrices and limits may change. Confirm the currency, billing period, seat minimum and usage caps on the official website.
Use Cases
- Model selection for products and apps: Compare leading LLMs on the leaderboard to choose the best model for chatbots, copilots, and content tools.
- Benchmarking in research and development: Use Arena rankings as a human-preference baseline to validate new models, fine-tunes, or system prompts.
- Market and competitive analysis: Track how open-source and commercial LLMs evolve over time to inform strategy, partnerships, or migration plans.
Features
Crowdsourced human model comparisons
Elo-style LLM performance ratings
Bradley–Terry statistical ranking
Side-by-side anonymous chat battles
Unified view of open and closed models
Continuously updated leaderboard data
Free web-based evaluation access
Granular insights across task types
Reviews
No reviews yet. Be the first to share your experience!
FAQ
What is LMSYS Chatbot Arena Leaderboard?
LMSYS Chatbot Arena Leaderboard is a free, crowdsourced platform that ranks large language models based on over one million human pairwise comparisons, using the Bradley–Terry model and Elo-style ratings to provide robust, real-world performance rankings.
How are the model scores and rankings calculated?
Scores are derived from human pairwise comparisons where users choose which of two anonymous models responds better. These outcomes are fed into the Bradley–Terry statistical model and converted to Elo-style ratings, producing a consistent ranking across many models and tasks.
Is LMSYS Chatbot Arena Leaderboard free to use?
Yes. The leaderboard is completely free to access in your browser. You can view rankings, inspect model details, and participate in chat battles without any subscription fees.
Can I use the leaderboard to choose a model for my product?
Yes. Many developers and teams use the leaderboard as a human-centric reference when selecting LLMs. You should still consider your specific domain, latency, cost, and safety needs, but Arena ratings provide a strong starting point.
Can I submit or benchmark my own model?
LMSYS periodically adds new models to the Arena. If you maintain an LLM and want it evaluated, you can follow the official documentation and submission channels on the LMSYS website or GitHub to request inclusion.