Skip to content
VECTOR WIREAI INTELLIGENCE
UTC

xbench-DeepSearch leaderboard — the family board.

2 TRACKED VARIANTS · 2 FIRST-CLASS BOARDS

One family, 2 tracked prints: xbench-DeepSearch (2025.10) is cited standalone and keeps live board of its own. Scoring configs and splits of the base board render as sections of this page, each under its own honest label.

● LIVE DATA14 MODELS ON THIS BOARDLEADER GPT-5 AT 77.8

First-class boards in this familyCITED STANDALONE · EACH KEEPS ITS OWN LIVE BOARD

#BoardModels scored
01xbench-DeepSearchTHIS PAGE14

The board — xbench-DeepSearchTOP 14 OF 14 MEASURED · BEST SCORE PER MODEL, DEFAULT VARIANT

#ModelVendorScoreVLast seen
01GPT-5OpenAI77.82026-06-06
02Kimi K2.5Moonshot76.72026-05-30
03Kimi K2 ThinkingMoonshot762026-05-30
04MiniMax M2MiniMax722026-06-06
05DeepSeek-V3.2DeepSeek712026-06-06
06GLM-4.6Z.ai702026-06-06
07Claude Sonnet 4.5Anthropic662026-06-06
08Claude Sonnet 4Anthropic64.62026-06-06
09Kimi K2 InstructMoonshot612026-06-06
10Step 3.5 FlashStepFun56.32026-05-30
11Gemini 2.5 ProGoogle562026-06-06
12GLM-4.7Z.ai52.32026-05-30
13MiMo V2 FlashXiaomi442026-05-30
14MiniMax M2.1MiniMax432026-05-30
◆ IV — INDEPENDENTLY VERIFIED · ◆ AA — AGGREGATOR · ◆ VA — VENDOR · ◇ SA — SOURCE-ATTRIBUTED
THE FAMILY BOARD · EVERY PRINT LABELED · EVERY SCORE SOURCE-ATTRIBUTED ON ITS OWN LIVE BOARDAll leaderboardsThe frontier boardModels by vendor