Agent Arena Index
AI overview
Sign in and the AI will write an overview from our coverage.
Headlines · 3
- Arena Launches Coding Index for AI Coding Agents
Arena introduced its Coding Index to evaluate coding agents using real coding work. Arena reported that GPT-6.2 Astra (Max) scored 1,338 points and ranked third, while its overall Agent Arena Index score was 1,291 points at sixth place. On the Coding Index, Arena said GPT-6.2 Astra led GPT-6 Astra (Max) by 50 points with a 62% lower median cost per task.
arena (X) · 🔥 8 - Gemini 4 Neon debuts with second-place Agent Arena ranking
Google DeepMind began rolling out Gemini 4 Neon, a model positioned between Gemini 4 Carbon and Gemini 4 Hydrogen for complex software engineering, knowledge work, and science. Arena reports that Gemini 4 Neon (High) ranked second overall on its Agent Arena Index with 1,336 points from 4.2k real-world agentic sessions, ahead of Carbon (1,303), Hydrogen (1,271), and Argon (1,229). Arena also places Neon on its cost-performance Pareto frontier at a median cost of $2.82 per task.
arena (X) · 🔥 5 - Arena Launches Agent Arena Index and Six Focused Indexes
Arena introduced the Agent Arena Index as its headline measure of overall agent performance, alongside six focused indexes: Agentic Work, Coding, Chat, Task Outcomes, Tool Use and Steering. Arena said the indexes are based on millions of real-world tasks and traces from its global community and combine signals including user-confirmed task completion, praise versus complaint, Bash reliability and steerability. Its initial Agent Arena Index leaderboard placed Claude Fable 5.2 (Max) first with 1,463 points, followed by Gemini 4 Carbon (High) with 1,307 and GPT-6 Astra (Max) with 1,274.
arena (X) · 🔥 5
Experience and discussion from the community
Share my Agent Arena Index experienceAsk about Agent Arena Index
Nobody has shared their experience with Agent Arena Index yet.