Your-Weights Leaderboard
Sliders for quality, speed, cost, context, safety.
Re-sort instantly and share the exact view via URL. The ranking follows your priorities—not a vendor's default formula.
ExploreAI Benchmark HubArena LeaderboardURL-shareable rankingsPublic API
Benchmark AI models by your quality.
A fast leaderboard and comparison hub that goes beyond one composite score: re-rank live, compare side-by-side, and stress-test in a multi-model arena.
5 dimensions·∞ rank orders·1 shareable URL
Three ways in
Everything you need to pick a model—without the spreadsheet.
Side-by-side Compare
Compare 2–4 models across 25+ fields.
Differences highlighted, plus export as Markdown.
ExploreLive Multi-Model Arena
Stream responses, see TTFT and tokens/sec.
Blind-vote the winner; votes feed community Elo.
ExploreComposite-free
No hidden default score
Instant resort
Weights update in real time
Community Elo
Arena votes shape the crowd rank
About AI Benchmark Hub
AI Benchmark Hub is the fastest way to choose an LLM for production: rank models with your own priorities, compare GPT vs Claude vs Gemini side-by-side, and stress-test prompts in a live arena at ai-benchmark-hub.com.
Your-weights LLM leaderboard
AI Benchmark Hub ranks frontier models using five dimensions you control: reasoning quality, inference speed, token cost, maximum context, and safety moderation. Drag sliders and the table re-sorts instantly—no spreadsheet required.
Open the model leaderboard →Side-by-side AI model comparison
Pick two to four models and compare pricing per million tokens, context limits, modalities, Arena Elo, and provider metadata. Differences are highlighted and you can export the table as Markdown for docs or procurement.
Compare models →Live multi-model AI arena
Stream answers from multiple LLMs on the same prompt, measure time-to-first-token and tokens per second, then blind-vote the winner. Community votes feed an Elo-style crowd ranking on AI Benchmark Hub.
Try the live arena →Popular searches
Frequently asked questions
What is AI Benchmark Hub?
AI Benchmark Hub is a free web app to rank, compare, and battle-test large language models (LLMs). Re-sort models by your own weights for quality, speed, cost, context window, and safety—then share the exact view via URL.
How is AI Benchmark Hub different from other AI leaderboards?
Unlike sites that hide everything behind one composite score, AI Benchmark Hub lets you set priorities with sliders, compare up to four models field-by-field, and run a live multi-model arena with blind voting and latency metrics.
Which AI models can I compare on AI Benchmark Hub?
The leaderboard and arena pull from the OpenRouter catalog—covering GPT, Claude, Gemini, Llama, Mistral, DeepSeek, Qwen, and hundreds more—with Arena Elo quality signals and live API pricing where available.
Is AI Benchmark Hub free to use?
Yes. Browsing the leaderboard, sharing weighted rankings, and comparing specs is free. The arena supports free-tier models; paid models use your own API keys when configured.
How do I compare GPT vs Claude?
Open our GPT vs Claude guide at /topics/gpt-vs-claude or go directly to Compare and pick OpenAI GPT and Anthropic Claude models. You will see pricing per million tokens, context limits, Arena Elo quality, and more.
What is the cheapest LLM API?
Cheapest models change often. On the leaderboard, set the cost weight to 100% to re-rank by blended price per million tokens using live OpenRouter data, or read /topics/cheapest-llm-api.
What is the best LLM for coding?
There is no single best coding LLM for every team. Rank models on the leaderboard with high quality and speed weights, compare finalists on /compare, then test real prompts in the /arena.
How does this compare to Chatbot Arena or LMSYS?
LMSYS Chatbot Arena focuses on crowd Elo from blind votes. AI Benchmark Hub adds live pricing, context, custom weighting, side-by-side specs, and a multi-model arena you control with your own prompts and API keys.