AGORA - Technology: leaderboard

Metric: Accuracy (%) on the 55 technology, software and manufacturing questions (547 documents), archive-grounded agentic document reasoning with verifiable numeric answers over the domain collection of Markdown workplace documents, agent in the mini-swe-agent harness (bash tool only), temperature 1.0 with reasoning effort at the maximum each model supports; higher is better. Source: arxiv.org. Saturation forecast: Around December 2026. 8 models tracked.

Top models

#ModelScore
1GPT-5.5 (xHigh)61.82
2GLM-5.152.73
3Gemini 3.1 Pro (Preview) (High)49.09
4DeepSeek V4 Pro (Max)47.27
5DeepSeek V4 Flash (Max)36.36
6Qwen 3.5 35B A3B16.36
7Gemini 3.1 Flash Lite (High)12.73
8Qwen 3.5 9B3.64

Interactive version: theaggregate.ai/benchmark?slug=agora-technology · How It Works · Data refreshed daily, snapshot 2026-09-29.