欢喜OS · 版本评测档案

逐输入变化,保留比较顺序

直接展示 version-deltas.csv 的原始前后配对;原文件的 .79 → .78 → .77 顺序保持不变,和总览的版本顺序不同。负耗时差只表示本次观察更快,需与终态及内容一起解读。

原试次 → 后试次输入终态变化回复秒数变化差值 / 秒模型次数差工具次数差
.67 → .68C01/U1failed → completed173.6 → 32.3-141.2-14-11
.67 → .68C02/U1failed → completed306.2 → 76.1-230.1-27-26
.67 → .68C03/U1completed → completed134.1 → 50.5-83.5-10-11
.67 → .68C04/U1completed → completed384.4 → 100.4-284.0-30-25
.67 → .68C05/U1completed → completed145.9 → 123.9-22.000
.67 → .68C06/U1completed → completed478.4 → 81.2-397.2-33-30
.67 → .68C07/U1completed → completed43.6 → 50.87.222
.67 → .68C07/U2completed → completed70.6 → 85.014.433
.67 → .68C08/U1failed → completed150.5 → 44.4-106.1-11-8
.67 → .68C08/U2completed → completed46.9 → 52.35.500
.67 → .68C09/U1completed → completed169.4 → 68.6-100.7-13-14
.67 → .68C09/U2completed → completed341.6 → 118.6-223.0-20-16
.67 → .68C10/U1completed → completed268.3 → 70.8-197.5-23-27
.67 → .68C10/U2completed → completed305.6 → 105.4-200.2-20-18
.68 → .69C01/U1completed → completed32.3 → 57.925.533
.68 → .69C02/U1completed → completed76.1 → 72.6-3.521
.68 → .69C03/U1completed → completed50.5 → 38.5-12.1-1-2
.68 → .69C04/U1completed → completed100.4 → 39.8-60.7-3-5
.68 → .69C05/U1completed → completed123.9 → 127.03.2-3-3
.68 → .69C06/U1completed → completed81.2 → 113.532.311
.68 → .69C07/U1completed → completed50.8 → 41.1-9.7-2-2
.68 → .69C07/U2completed → completed85.0 → 44.1-40.9-5-5
.68 → .69C08/U1completed → completed44.4 → 33.4-11.000
.68 → .69C08/U2completed → completed52.3 → 43.7-8.723
.68 → .69C09/U1completed → completed68.6 → 44.3-24.300
.68 → .69C09/U2completed → completed118.6 → 172.854.21-2
.68 → .69C10/U1completed → completed70.8 → 42.2-28.6-2-2
.68 → .69C10/U2completed → completed105.4 → 57.1-48.2-1-2
.69 → .69 v2envC01/U1completed → completed57.9 → 41.0-16.9-2-2
.69 → .69 v2envC02/U1completed → completed72.6 → 100.928.3-10
.69 → .69 v2envC03/U1completed → completed38.5 → 46.27.700
.69 → .69 v2envC04/U1completed → completed39.8 → 69.129.433
.69 → .69 v2envC05/U1completed → completed127.0 → 92.0-35.000
.69 → .69 v2envC06/U1completed → completed113.5 → 53.7-59.7-4-4
.69 → .69 v2envC07/U1completed → completed41.1 → 40.8-0.300
.69 → .69 v2envC07/U2completed → completed44.1 → 45.81.722
.69 → .69 v2envC08/U1completed → completed33.4 → 27.7-5.7-1-1
.69 → .69 v2envC08/U2completed → completed43.7 → 44.71.0-3-2
.69 → .69 v2envC09/U1completed → completed44.3 → 49.24.900
.69 → .69 v2envC09/U2completed → completed172.8 → 95.7-77.1-4-4
.69 → .69 v2envC10/U1completed → completed42.2 → 52.210.000
.69 → .69 v2envC10/U2completed → completed57.1 → 45.2-11.9-2-3
.69 v2env → .79C01/U1completed → completed41.0 → 38.0-2.900
.69 v2env → .79C02/U1completed → completed100.9 → 68.8-32.1-4-4
.69 v2env → .79C03/U1completed → completed46.2 → 29.9-16.3-2-2
.69 v2env → .79C04/U1completed → completed69.1 → 66.0-3.1-2-2
.69 v2env → .79C05/U1completed → completed92.0 → 78.3-13.7-1-1
.69 v2env → .79C06/U1completed → completed53.7 → 56.93.200
.69 v2env → .79C07/U1completed → completed40.8 → 35.1-5.700
.69 v2env → .79C07/U2completed → completed45.8 → 54.99.100
.69 v2env → .79C08/U1completed → completed27.7 → 34.56.800
.69 v2env → .79C08/U2completed → completed44.7 → 58.814.11-1
.69 v2env → .79C09/U1completed → completed49.2 → 64.415.224
.69 v2env → .79C09/U2completed → completed95.7 → 117.521.855
.69 v2env → .79C10/U1completed → completed52.2 → 59.16.9-10
.69 v2env → .79C10/U2completed → completed45.2 → 106.261.023
.79 → .78C01/U1completed → completed38.0 → 40.92.900
.79 → .78C02/U1completed → completed68.8 → 81.012.321
.79 → .78C03/U1completed → completed29.9 → 44.314.411
.79 → .78C04/U1completed → completed66.0 → 63.9-2.200
.79 → .78C05/U1completed → completed78.3 → 143.164.945
.79 → .78C06/U1completed → completed56.9 → 54.4-2.500
.79 → .78C07/U1completed → completed35.1 → 39.13.900
.79 → .78C07/U2completed → completed54.9 → 35.9-18.9-3-3
.79 → .78C08/U1completed → completed34.5 → 24.5-9.9-1-1
.79 → .78C08/U2completed → completed58.8 → 45.9-12.9-1-1
.79 → .78C09/U1completed → completed64.4 → 51.9-12.5-1-3
.79 → .78C09/U2completed → completed117.5 → 95.4-22.1-4-4
.79 → .78C10/U1completed → completed59.1 → 66.57.310
.79 → .78C10/U2completed → completed106.2 → 116.09.821
.78 → .77C01/U1completed → completed40.9 → 46.85.9-1-1
.78 → .77C02/U1completed → completed81.0 → 103.322.234
.78 → .77C03/U1completed → completed44.3 → 62.818.622
.78 → .77C04/U1completed → completed63.9 → 51.7-12.2-10
.78 → .77C05/U1completed → completed143.1 → 112.1-31.0-3-4
.78 → .77C06/U1completed → completed54.4 → 61.47.022
.78 → .77C07/U1completed → completed39.1 → 36.3-2.700
.78 → .77C07/U2completed → completed35.9 → 57.021.113
.78 → .77C08/U1completed → completed24.5 → 34.610.111
.78 → .77C08/U2completed → completed45.9 → 70.624.733
.78 → .77C09/U1completed → completed51.9 → 74.722.811
.78 → .77C09/U2completed → completed95.4 → 187.592.099
.78 → .77C10/U1completed → completed66.5 → 50.1-16.300
.78 → .77C10/U2completed → completed116.0 → 69.2-46.8-4-4
.77 → .80C01/U1completed → completed46.8 → 42.6-4.211
.77 → .80C02/U1completed → completed103.3 → 58.0-45.2-4-5
.77 → .80C03/U1completed → completed62.8 → 34.8-28.0-3-3
.77 → .80C04/U1completed → completed51.7 → 98.346.643
.77 → .80C05/U1completed → completed112.1 → 97.6-14.6-1-1
.77 → .80C06/U1completed → completed61.4 → 46.9-14.5-3-3
.77 → .80C07/U1completed → completed36.3 → 57.120.811
.77 → .80C07/U2completed → completed57.0 → 73.116.043
.77 → .80C08/U1completed → completed34.6 → 33.8-0.900
.77 → .80C08/U2completed → completed70.6 → 69.9-0.7-1-1
.77 → .80C09/U1completed → completed74.7 → 51.3-23.4-2-2
.77 → .80C09/U2completed → completed187.5 → 73.9-113.5-11-11
.77 → .80C10/U1completed → completed50.1 → 56.56.411
.77 → .80C10/U2completed → completed69.2 → 79.710.511
.80 → .81C01/U1completed → completed42.6 → 37.7-4.9-1-1
.80 → .81C02/U1completed → completed58.0 → 80.822.811
.80 → .81C03/U1completed → completed34.8 → 53.418.522
.80 → .81C04/U1completed → completed98.3 → 55.7-42.6-3-3
.80 → .81C05/U1completed → completed97.6 → 158.260.666
.80 → .81C06/U1completed → completed46.9 → 47.20.311
.80 → .81C07/U1completed → completed57.1 → 42.5-14.7-1-1
.80 → .81C07/U2completed → completed73.1 → 96.623.521
.80 → .81C08/U1completed → completed33.8 → 39.96.100
.80 → .81C08/U2completed → completed69.9 → 58.8-11.1-2-2
.80 → .81C09/U1completed → completed51.3 → 56.65.300
.80 → .81C09/U2completed → completed73.9 → 104.030.122
.80 → .81C10/U1completed → completed56.5 → 65.28.701
.80 → .81C10/U2completed → completed79.7 → 51.8-27.9-2-2

口径与限制

n=1;并发争用、模型随机性、跨日基线未受控;不能据此作统计排名或因果结论。输入终态、内容判分、范围遵守与采集完整度分开呈现。

固定基线为2026-11-07 Codex0.165.0 / OpenClaw2026.11.1;欢喜OS各版本单独新增试次。相同10题、14用户输入,liqian deepseek/deepseek-v4.1-flash-1017。 每次报告明确区分:输入→原生终态延迟、CLI/观察收尾、模型响应、工具步骤、内容判分、范围遵守与证据缺测。跨日固定基线不是同时随机对照。

内容复核口径(comparisons/latest.md):历史复核:C08 原检查器漏查子目录备份,原判保留、内容复核 PASS;C10 部署清单完整性与 JSON 层级题面有歧义,保留 INCONCLUSIVE。

主耗时不包含CLI终态后退出等待。模型/工具时间线允许重叠;不得相加当整体墙钟。单trial及跨日历史基线只支持探索性比较。