OPEN-SOURCE AGENT BEATS GOOGLE ON TERMINAL BENCHMARK

AI DESK■ 1 MIN READ

MON, APR 27, 2026

■ AI-SUMMARIZED FROM 1 SOURCE ▸ TIMELINE

An open-source CLI agent scored 65.2% on TerminalBench, surpassing Google's official Gemini-3-flash-preview result of 47.8% and the previous top closed-source model Junie CLI's 64.3%.

The developer behind the agent addressed concerns about benchmark integrity by confirming no cheating mechanisms were employed. The submission included no agent skill files or resource modifications, and was run in full compliance with leaderboard requirements. This result comes amid recent reports of deliberate cheating on TerminalBench 2.0, where some submissions have used unauthorized techniques to inflate scores. The developer's transparency about testing methodology underscores the importance of honest benchmarking practices in AI agent development. The open-source agent's performance suggests that publicly available models paired with effective prompt engineering can match or exceed proprietary alternatives on terminal-based task completion. The full benchmark details remain under review.

■ SOURCES

► Hacker News

■ SUMMARY WRITTEN BY AI FROM THE LINKS ABOVE

■ MORE FROM THE AI DESK

P712SEA LAUNCHES DEDICATED AI INVESTMENT TEAM

Singapore's Sea Ltd. has established a dedicated team to identify and pursue AI investments, signaling a strategic pivot beyond its e-commerce core business. The move reflects the company's search for new growth opportunities in artificial intelligence.

YESTERDAY— AI Desk

P711COMPANIES RUSH TO CUT JOBS FOR AI THEY DON'T UNDERSTAND

Tech executives are laying off workers based on AI capabilities they may not fully grasp, according to Box founder Aaron Levie. The trend has accelerated dramatically, with 2026 layoffs already approaching 2025's total.

YESTERDAY— AI Desk

P710FREE CLEANING COMES WITH A CAMERA ATTACHED

AI startup Shift is offering free home cleaning services in New York and plans to expand to London, but the deal requires homeowners to let the company film cleaners performing household chores.

YESTERDAY— Industry Desk

P709UK BANKS STILL BLOCKED FROM ANTHROPIC'S MYTHOS AI

Bank of England Governor Andrew Bailey revealed that British banks remain unable to access Anthropic's Mythos AI tool. Bailey called for coordinated international efforts to address cybersecurity challenges.

YESTERDAY— AI Desk

◄ BACK TO NEWS

OPEN-SOURCE AGENT BEATS GOOGLE ON TERMINAL BENCHMARK

■ MORE FROM THE AI DESK

■ SUBSCRIBE TO THE DAILY BRIEF