Loading activity…
@arena
Where AI meets the real world. We measure and advance the frontier of AI through community-driven evaluation. We’re hiring → http://arena.ai/jobs
AI
Agent Arena says GPT-6 Luna takes three of four optimal spots below $0.10 per task at three reasoning levels.
AI
Artificial Analysis put Ultra fifth on AA-Image-Editing v2.1; Arena ranked it fourth in Multi-Image Edit on October 20.
AI
Arena places the model 11th overall in Agent Arena, with a 5.1% net improvement across 7,500-plus real-world sessions.
Technology
MiniMax says the model is nearly three times faster and 70% cheaper for agentic tasks than its predecessor.
AI
A commentator suggests MiniMax is also using smaller models and links the shift to compute scarcity and reinforcement learning.
AI
Arena says the model scored 1,670 points, ranking 12th overall and 52 points above the previous version.
AI
Arena says the index is built from 90K-plus real-world agent sessions across 30 models.
Technology
The company says the model improves benchmarks, task speed and output length over its predecessor.
AI
DeepSeek says the model is live on its website, app and API, and optimized for DeepSeekHarness and Tencent CodeBuddy.
AI
Agent Arena puts OpenAI and Anthropic 35 points above its leaderboard average on praise versus complaints.
AI
Reflection says the release includes Mirror CLI, its open-source training harness, and collaborations with 40-plus partners.
AI
Agent Arena says Pro’s +3.8% net improvement puts it ahead of second-place Deepseek V4.1 Flash (Max) at +3.0%.
Technology
Meta says its 816B-parameter model can run locally across two M5 Ultra Mac Studios.
AI
finkd says the model runs well on new Mac Studios, crediting work with Apple and John Ternus.
AI
Arena says Max Spero covers synthetic mirrors, model-family fingerprints and AI-assisted writing in a conversation about Pangram’s training process.
AI
Arena says Muse Wonder (Max) climbed nine places from Muse Spark 1.3 (Max)’s ranking in Text Arena.
AI
Agent Arena says its Bash Reliability signal evolved from Bash Recovery, while proprietary models occupy the top 14 spots.
AI
The Arena says Sonnet sits nine points below Opus at half the blended input/output token price.
AI
Arena says it tracks five signals from real-world, long-horizon agent sessions instead of using pairwise preference votes.
AI
Arena says its index tracks unauthorized actions, false attribution and deceptive completion across 90,000-plus agent sessions and 27 models.
YouTube
