Company intelligence
Arena
Where AI meets the real world.
About Arena
Created by researchers from UC Berkeley, Arena (formerly LMArena) is a community-powered platform to measure and advance the frontier of AI for real-world use. Tens of millions of builders, researchers, and creative professionals come to Arena to use frontier models and give feedback on their responses, shaping a public leaderboard grounded in real-world use.
Verified activity
Signals from Arena
12 published signals
Products & Services
Arena reported that GPT-6 Astra (Max) scored a massive 40.6% increase in Praise vs. Complaint and 21.9% increase in Confirmed Success.
Reported by Arena
Products & Services
Arena noted that GPT-6 Astra (Max) costs per median task are $4.01, reshaping the Pareto frontier for Agent Arena.
Reported by Arena
Products & Services
Arena congratulated OpenAI on the release of GPT-6 Astra (Max) debuting at #2 overall with a 12.5% net improvement across 8.9K real-world agentic sessions.
Reported by Arena
Products & Services
Arena announced that GPT-6 Astra (Max) debuts at #2 overall with a 12.5% net improvement across 8.9K real-world agentic sessions.
Reported by Arena
Products & Services
Arena announced that Claude Fable 5.1 (Max) by Anthropic has landed in the Agent Arena at #1, achieving a $4.14 median cost per task and a +15.8% net improvement.
Reported by Arena
Presence & Recognition
Arena measured Claude Fable 5.1 (Max) by Anthropic in the Agent Arena at #1 with a +15.8% net improvement across 6.7k+ real-world agentic sessions.
Reported by Arena
Presence & Recognition
Arena announced that the Grok Imagine Video 1.5 Agent from SpaceXAI landed in the Text-to-Video Arena at position 5 with 1491 points.
Reported by Arena
Presence & Recognition
Arena announced that GPT-6 Astra (Max) by OpenAI is ranked #1 on Code Arena: WebDev, achieving SOTA performance at $40/Mtoken and matching the latest Claude model pricing.
Reported by Arena
Presence & Recognition
Arena achieved a new #1 position on Code Arena with GPT-6 Astra (Max) scoring 1797 points.
Reported by Arena
Products & Services
Arena engineers Graham Tremper and Logan King built the same taco truck waitlist app from a single napkin sketch using Agent Mode.
Reported by Arena
Research & Knowledge
Arena published a comparison of GPT-6 Astra vs. Claude Fable 5.1 vs. Claude Fable 5 on cost, coding, and benchmarks.
Reported by Arena
People
Arena is hiring AI Infrastructure and Backend Engineers to build the core infrastructure powering the next generation of AI products.
Reported by Neng Lu