Will any AI model reach 1600 Coding Arena Score by December 31, 2026?
As of September 4, 2026, the market gives “Will any AI model reach 1600 Coding Arena Score by December 31, 2026?” a 86% chance of NO.
| YES odds | 14% |
|---|---|
| NO odds | 86% |
| Volume | $35,094 |
| Closes | January 1, 2027 |
Updated September 4, 2026 · Live data from Polymarket
About this market
Recent model releases from Anthropic and OpenAI continue to drive trader sentiment on whether any AI reaches the target Coding Arena score by year-end. Claude Fable 5.1 and Opus 5 variants lead multiple coding arenas and agent benchmarks, including WebDev Arena Elo ratings near 1765 and strong results on SWE-bench Verified (up to 96%) and FrontierSWE, reflecting advances in agentic workflows and long-horizon software engineering. OpenAI’s GPT-6 Astra and GPT-5.6 series sit close behind on several leaderboards, while Qwen and GLM models show competitive Elo in preference-based arenas. With four months remaining, further capability jumps from scaling, harness improvements, or new releases before December 31 could shift outcomes, though benchmark saturation on easier tasks and slower progress on harder agentic ones create ongoing uncertainty in trader consensus.