pingdotgg/t3code

Run by mupt-ai · Released 2026-10-02 · 39 tasks · 9 settings

Model Tradeoffs

Which coding agent works best on pingdotgg/t3code? Most accurate: Claude Opus 5.5, 76.9% at $2.05 per task. Best for less: GPT-6.1 Sol, 66.7% at $0.31. 9 model settings scored on 39 tasks from its merged pull requests.

All Settings

ModelHarnessReasoningAccessAccuracyCost / TaskFrontier
Claude Opus 5.5Claude CodeHighAPI Key76.9%$2.05Yes
GPT-6.1 SolCodexLowAPI Key66.7%$0.31Yes
GLM 5.3 FlashPiHighVercel AI Gateway64.1%$0.35No
GPT-6.1 SolCodexHighAPI Key64.1%$0.60No
Claude Sonnet 5.5Claude CodeHighAPI Key59.0%$0.87No
Kimi K3PiHighVercel AI Gateway56.4%$2.38No
GLM 5.3CodexHighVercel AI Gateway51.3%$3.01No
GPT-6 LunaCodexHighAPI Key46.2%$0.037Yes
GPT-6 LunaCodexLowAPI Key25.6%$0.041No

Tasks

All 39 tasks, each with its instruction, tests, and solution to browse and download.

All Repositories on SelfBench