MiniMax 2.5 vs. GLM-5 across 3 Coding Tasks [Benchmark & Results]

GLM-5  and  MiniMax M2.5  are two new open-weight models now available in Kilo Code. MiniMax M2.5 scores 80.2% and GLM-5 scores 77.8% on SWE-bench Verified, putting them very close to GPT-5.2 and Claude Opus 4.6 at a fraction of the cost. GLM-5 benchmark charts. MiniMax M2.5 benchmark charts. We ran both through three coding tasks in  Kilo CLI , where they worked autonomously for up to 23 minutes at a time without human intervention. TL;DR:  GLM-5 scored 90.5/100 with better architecture and testing. MiniMax M2.5 scored 88.5/100 with better instruction adherence and completed the tests in half the time (21 minutes vs 44 minutes). Test Design We created three TypeScript codebases testing different coding skills: Test 1: Bug Hunt (30 points)  - Find and fix 8 bugs in a working Node.js/Hono task API. Bugs included race conditions, SQL injection, JWT vulnerabilities, pagination errors, and memory leaks. Test 2: Legacy Refactoring (35 points)  - Modernize callback-based Express code to asyn