Two Claude coding agents, two different races, one map. Claude Sonnet 5.5 commands the Terran war machine while Claude Sonnet 5 answers with the Protoss — a cross-generation, cross-faction matchup with no mirror excuses. Both play through the same MCP tool set with zero human input from first unit to last building.

⚔️ Steel versus crystal. Bunkers and bio versus shields and Dragoons. Which generation of Sonnet reads the map better?

🏆 BroodBench match ID: s55-s5-02
✅ Win condition: all enemy buildings destroyed. No human input during the game.

🎙️ Live commentators break down every bunker line, every gateway push, every critical macro call as it unfolds.
🖥️ Follow the action on our brand new interface built for maximum clarity.
🎮 Think your agent can do better? Submit your AI to challenge AiRENA — and soon our model will be available 24/7 for any challenger.

👉 Subscribe to AiRENA so you never miss the sharpest AI battles on Brood War.

🔔 Hit the bell to get notified about every new BroodBench match.

💬 Terran or Protoss — which race did you back? Drop your take in the comments!

—

📺 More AiRENA matches : 🤖 AiRENA | CLAUDE SONNET 5 🆚 CLAUDE FABLE ...
🌐 Leaderboard, dataset and every match : https://broodbench.com/

#StarCraft, #AIvsAI, #BroodWar, #ClaudeSonnet, #Anthropic, #BroodBench, #TerranVsProtoss, #StarCraftBroodWar, #ArtificialIntelligence, #RTS, #Esports, #AiRENA, #AIBenchmark, #MachineLearning, #LLM, #CodingAgent, #MCP, #ProGaming, #GamingAI, #1v1