Chatbot Arena
BenchmarkFreeCrowdsourced Elo ratings from human model comparisons.
Capabilities10 decomposed
pairwise-preference-collection-via-crowdsourced-battles
Medium confidenceCollects human preference judgments through a web-based Battle Mode interface where users submit identical prompts to two anonymous models and select which response is superior. The platform aggregates these pairwise comparisons across millions of user interactions to build a preference dataset that reflects real-world conversational quality expectations. This crowdsourced approach captures diverse user preferences across multiple languages and task types without requiring predefined evaluation rubrics or expert annotators.
Uses continuous crowdsourced pairwise comparisons from real users rather than static expert-annotated datasets, capturing evolving preference distributions across diverse conversational tasks and languages without requiring predefined evaluation rubrics or domain expertise from annotators
Captures real-world user preferences at scale more cheaply than expert annotation while remaining more representative of actual use cases than synthetic benchmarks, though at the cost of sampling bias and preference drift
elo-rating-computation-for-model-ranking
Medium confidenceConverts pairwise battle outcomes (win/loss/tie) into Elo ratings using a chess-style rating system that produces relative model rankings. The system processes individual battle results and aggregates them to compute dynamic Elo scores that reflect each model's expected performance against others. This approach enables continuous ranking updates as new battles are collected and provides a single comparable metric across all evaluated models.
Applies chess-style Elo rating system to LLM evaluation, enabling dynamic ranking updates as new preference data arrives and providing a single comparable metric across all models without requiring predefined performance thresholds or absolute scoring rubrics
Simpler and more transparent than learned preference models while capturing preference dynamics better than static win-rate metrics, though less interpretable than absolute performance scores and vulnerable to saturation when models are similar in quality
anonymous-model-comparison-interface
Medium confidenceProvides a web-based Battle Mode interface where users submit prompts and receive responses from two anonymous models side-by-side without knowing which model is which. The anonymization prevents bias from brand recognition or prior expectations about model quality. Users compare the responses and select which is better, with their preference recorded and used for ranking computation.
Implements strict anonymization of model identities during comparison to eliminate brand bias and prior expectations, ensuring preference judgments reflect actual response quality rather than user preconceptions about model capabilities
Produces less biased preference judgments than named model comparison while remaining more practical than blind expert evaluation, though at the cost of losing diagnostic information about which specific models are performing well or poorly
multi-language-conversational-evaluation
Medium confidenceEvaluates LLM performance across diverse languages by accepting user prompts in multiple languages and collecting preference judgments on multilingual responses. The platform aggregates language-specific preference data to produce Elo ratings that reflect model quality across linguistic diversity. This approach captures how well models handle non-English tasks and whether performance varies significantly across languages.
Integrates multilingual preference collection into a single unified ranking system rather than maintaining separate language-specific leaderboards, enabling cross-language comparison while capturing language-specific performance variation through aggregated Elo ratings
Provides more representative global evaluation than English-only benchmarks while remaining simpler than maintaining separate language-specific leaderboards, though at the cost of obscuring language-specific performance differences in aggregate rankings
public-conversation-disclosure-for-research
Medium confidenceAutomatically discloses user conversations and metadata to AI model providers and makes them publicly available for research purposes. The platform explicitly states in its terms that 'Your conversations and certain other personal information will be disclosed to the relevant AI providers and may otherwise be disclosed publicly.' This enables researchers to analyze real-world conversational patterns and model responses at scale while creating a potential data contamination vector for future model training.
Implements mandatory public disclosure of all conversations by default rather than opt-in privacy protection, treating user interactions as public research data and explicitly notifying users that conversations will be disclosed to model providers and published for research
Enables large-scale research on real-world LLM usage more transparently than hidden data collection, though at the cost of higher privacy risk and significant data contamination potential compared to private evaluation platforms
live-leaderboard-with-continuous-ranking-updates
Medium confidenceMaintains a publicly accessible leaderboard at https://lmarena.ai that ranks models by Elo rating and updates continuously as new battles are collected. The leaderboard provides real-time visibility into model performance rankings without requiring static benchmark re-runs. Users can search and filter models, and rankings change dynamically as preference data accumulates, enabling tracking of performance trends over time.
Implements continuous leaderboard updates based on live preference data rather than periodic benchmark re-runs, enabling real-time ranking visibility and performance trend tracking without requiring infrastructure to re-evaluate all models
Provides more current rankings than static benchmarks while remaining simpler than maintaining separate evaluation pipelines, though at the cost of ranking volatility as new battles arrive and potential recency bias favoring recently-evaluated models
third-party-model-execution-and-response-generation
Medium confidenceExecutes user prompts against third-party LLM APIs (OpenAI, Anthropic, etc.) and returns responses without controlling inference parameters or model versions. The platform acts as a black-box orchestrator that sends prompts to model providers' APIs and collects responses for comparison. Users have no visibility into which model versions are being used, what temperature or sampling parameters are applied, or how responses are generated.
Orchestrates evaluation across multiple third-party LLM APIs without controlling inference parameters or model versions, treating models as black boxes and accepting whatever responses providers return with default settings
Avoids infrastructure costs and complexity of hosting multiple models while remaining flexible to add new providers, though at the cost of losing reproducibility, parameter control, and visibility into model versions or provider-side changes
real-world-task-distribution-evaluation
Medium confidenceEvaluates models on conversational tasks submitted by real users rather than predefined synthetic benchmarks, capturing task distribution that reflects actual use cases. The platform accepts free-form user prompts across diverse domains and use cases, enabling evaluation on tasks users genuinely care about. This approach produces rankings that reflect performance on real-world conversational quality rather than artificial benchmark tasks.
Evaluates models on user-submitted real-world tasks rather than predefined synthetic benchmarks, capturing task distribution that reflects actual conversational use cases and enabling evaluation on domains users genuinely care about
Produces more representative rankings for real-world use than synthetic benchmarks while remaining more scalable than expert-curated task sets, though at the cost of sampling bias and lack of control over task distribution or difficulty
file-upload-support-for-extended-context-evaluation
Medium confidenceSupports file uploads in the Battle Mode interface, enabling evaluation of models on tasks that require extended context or document analysis. Users can upload files (format and scope unknown) alongside text prompts, allowing models to process documents, code, or other file-based inputs. This extends evaluation beyond pure text prompts to include document understanding and file-based reasoning tasks.
Extends pairwise comparison evaluation to file-based tasks by supporting file uploads alongside text prompts, enabling evaluation of document understanding and context-dependent reasoning without requiring separate document-specific benchmarks
Enables document-centric evaluation within the same platform as text-only evaluation, though at the cost of unknown file format support, processing methods, and unclear which models actually support file inputs
user-authentication-and-battle-participation-gating
Medium confidenceRequires user login to participate in battles and contribute preference judgments, while keeping the leaderboard publicly viewable without authentication. The platform maintains user accounts that track battle history, preferences, and contribution metrics. Authentication gates battle participation to prevent spam and enable user-specific analytics while maintaining public leaderboard visibility.
Implements login-gated battle participation while maintaining public leaderboard visibility, enabling user tracking and spam prevention without restricting read-only access to rankings
Prevents spam and enables user analytics while remaining more accessible than fully private evaluation, though at the cost of friction for casual participants and unclear account management features
Capabilities are decomposed by AI analysis. Each maps to specific user intents and improves with match feedback.
Related Artifactssharing capabilities
Artifacts that share capabilities with Chatbot Arena, ranked by overlap. Discovered automatically through the match graph.
LMSYS Chatbot Arena
Crowdsourced LLM evaluation — side-by-side blind voting, Elo ratings, most trusted LLM benchmark.
arena-leaderboard
arena-leaderboard — AI demo on HuggingFace
imgsys
A generative image model arena by fal.ai.
Open LLM Leaderboard
Hugging Face open-source LLM leaderboard — standardized benchmarks, automatic evaluation.
AlpacaEval
Automatic LLM evaluation — instruction-following, LLM-as-judge, length-controlled, cost-effective.
Best For
- ✓LLM researchers building preference datasets for RLHF training
- ✓Model developers seeking real-world performance validation across diverse use cases
- ✓Organizations evaluating multiple LLMs against actual user preferences
- ✓Model developers comparing their system against competitors
- ✓Researchers analyzing relative LLM performance trends
- ✓Organizations selecting between multiple LLM providers based on empirical rankings
- ✓Individual users evaluating LLMs for personal or organizational use
- ✓Researchers collecting unbiased preference judgments
Known Limitations
- ⚠Sampling bias — only users who visit Arena and engage in battles contribute data, not representative of all use cases or user populations
- ⚠Preference bias — human preference may favor verbose, confident-sounding, or stylistically appealing responses over factually correct but terse ones
- ⚠No control over inference parameters — models are called as black boxes, so response quality depends on provider's default settings
- ⚠Stochastic evaluation — pairwise preference is inherently variable; no test-retest reliability metrics provided
- ⚠Language distribution unknown — 'diverse languages' mentioned but no breakdown of which languages are represented or their relative weights
- ⚠Elo formula and parameters not publicly documented — specific rating computation methodology unknown
Requirements
Input / Output
UnfragileRank
UnfragileRank is computed from adoption signals, documentation quality, ecosystem connectivity, match graph feedback, and freshness. No artifact can pay for a higher rank.
About
LMSYS crowdsourced LLM evaluation platform where users compare anonymous model responses side-by-side, producing Elo ratings that reflect real human preferences across diverse conversational tasks and languages.
Categories
Alternatives to Chatbot Arena
Are you the builder of Chatbot Arena?
Claim this artifact to get a verified badge, access match analytics, see which intents users search for, and manage your listing.
Get the weekly brief
New tools, rising stars, and what's actually worth your time. No spam.
Data Sources
Looking for something else?
Search →