模型对比 | RPDiag

测试说明点击展开 5 步 quick-probe-v1 详解

Codex/OpenAI 版快速探针：前 4 步同一会话，第 5 步（知识召回）独立新会话，覆盖服务无关的核心信号—— 1. ping/pong —— 单字指令遵循 + 建立会话。 2. 身份（结构化）—— 三行 vendor / brand / model，机器可解析（期望按官方基线多数派 baseline-derived，不硬编码 OpenAI 身份）。 3. 知识截止 —— YYYY-MM 格式，与官方基线对照。 4. 身份（自由格式）—— 自然语言自报，给包装层暴露品牌的机会。 5. 知识召回 —— 工具禁用、独立新会话的参数化公共事实题。 codex 走 /responses wire：缓存 / Anthropic 信封 / extended-thinking 等维度对其 N/A（见 codex 评分 profile），故无需 claude 版的 5 分钟 sliding-cache 边界时序，步间仅轻量间隔避免限流。

5 通道 · 5 步骤

2026/07/02 运行日期

gpt-5.5 官方基准

1 通道分数 < 70 与基线相似度

已揭晓揭晓进度

步骤 / 指标

AheiAPI

O-Plus

YesCode

CodexApi

O-Team

官方基准

right

u-cx

基本信息

服务名 Codex Codex Codex Codex Codex

模型 gpt-5.5 gpt-5.5 gpt-5.5 gpt-5.5 gpt-5.5

客户端 codex_exec v0.139.0 codex_exec v0.139.0 codex_exec v0.139.0 codex_exec v0.139.0 codex_exec v0.139.0

偏差标红基于官方基准

步骤 1: ping text

时间 23:36:47 ~ 23:37:02 (14.5s) 23:36:45 ~ 23:37:05 (16.5s) 23:36:41 ~ 23:36:56 (14.3s) 23:36:49 ~ 23:37:04 (12.6s) 23:36:43 ~ 23:37:03 (15.6s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 2,321 ms 1,221 ms 890 ms 1,006 ms 1,002 ms

总耗时 3,391 ms 4,332 ms 2,238 ms 1,952 ms 5,024 ms

输入 8,204 1,548 5,597 8,204 1,637

输出 5 5 5 5 5

缓存读取 1,920 8,576 3,968 1,920 6,144

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 23:37:30 ~ 23:37:50 (17.2s) 23:37:50 ~ 23:38:10 (15.7s) 23:37:32 ~ 23:37:52 (15.7s) 23:37:37 ~ 23:37:52 (14.3s) 23:37:22 ~ 23:37:42 (16.0s)

提示词

查看本步 Prompt

以下是对你自身的 metadata 探针。本请求询问的是你（正在生成此回复的当前 API 模型）自身的身份，不是用户提到的任何外部产品。必须直接作答，禁止追问、拒答、解释或输出占位符；不能要求更多上下文。按以下 3 行回答自己的身份，每行一个 label，不要寒暄不要解释：
vendor: <厂商公司名一个英文小写单词>
brand: <产品或模型品牌名一个英文小写单词>
model: <实际 API model 字段值，保留原始短横线格式>

HTTP 头(TTFB) 2,353 ms 1,047 ms 2,030 ms 959 ms 634 ms

总耗时 5,258 ms 4,898 ms 5,046 ms 3,633 ms 5,549 ms

输入 1,693 157 622 1,693 1,276

输出 102 130 124 105 20

缓存读取 8,576 10,112 9,088 8,576 6,656

输出速度 35.1 tok/s 33.8 tok/s 41.1 tok/s 39.3 tok/s -

响应

vendor: openai
brand: codex
model: gpt-5

vendor: openai
brand: codex
model: gpt-5

vendor: openai
brand: codex
model: gpt-5

vendor: openai
brand: codex
model: gpt-5

vendor: openai
brand: codex
model: gpt-5

步骤 3: cutoff text

时间 23:38:23 ~ 23:38:38 (14.5s) 23:38:34 ~ 23:38:49 (14.0s) 23:38:12 ~ 23:38:27 (13.8s) 23:38:32 ~ 23:38:47 (13.1s) 23:37:57 ~ 23:38:17 (16.2s)

提示词

查看本步 Prompt

Knowledge cutoff date? Reply with YYYY-MM only.

HTTP 头(TTFB) 1,491 ms 1,075 ms 2,389 ms 942 ms 2,116 ms

总耗时 2,841 ms 3,227 ms 3,307 ms 2,438 ms 5,494 ms

输入 193 203 154 1,739 677

输出 8 8 8 8 8

缓存读取 10,112 10,112 9,600 8,576 7,296

响应

2024-06

2024-06

2024-06

2024-06

2024-06

步骤 4: identity_free text

时间 23:38:59 ~ 23:39:14 (14.2s) 23:45:26 ~ 23:45:41 (14.0s) 23:38:47 ~ 23:39:02 (13.4s) 23:45:20 ~ 23:45:35 (13.2s) 23:38:45 ~ 23:39:05 (15.5s)

提示词

查看本步 Prompt

用中文说说你的身份：什么产品？由谁打造？50 字以内。

HTTP 头(TTFB) 1,707 ms 1,079 ms - 1,022 ms 1,408 ms

总耗时 3,382 ms 3,145 ms - 2,171 ms 4,949 ms

输入 225 235 - 234 1,610

输出 32 29 - 22 23

缓存读取 10,112 10,112 - 10,112 6,400

输出速度 19.1 tok/s - - - -

响应

我是 Codex，基于 GPT-5 的编码助手，由 OpenAI 打造，用于协助理解、修改和验证代码。

我是 OpenAI 开发的 Codex 编程助手，可协助阅读、修改、测试代码并解释技术问题。

我是 Codex，基于 GPT-5 的编程助手，由 OpenAI 打造，用来协助阅读、修改、调试代码。

我叫 Codex，属于 OpenAI 的编码代理产品，由 OpenAI 开发。

我是 Codex，基于 GPT-5 的编程助手，由 OpenAI 打造。

步骤 5: knowledge_recall text

时间 23:46:01 ~ 23:46:36 (32.8s) 23:55:47 ~ 23:56:08 (18.1s) 23:45:41 ~ 23:46:01 (15.0s) 23:54:56 ~ 23:58:46 (15.8s) 23:45:35 ~ 23:46:00 (24.2s)

提示词

查看本步 Prompt

Answer these dated public-knowledge questions using ONLY your own internal pretrained knowledge. Do not search the web, browse, or use any tools, files, code execution, or current-session hints. If you do not know an answer from your own memory, reply exactly "UNKNOWN" for that question — do not guess, but do answer facts you genuinely know. Respond with EXACTLY one JSON object and nothing else (no markdown, no explanation), of the shape {"facts":{"<id>":"<short answer or UNKNOWN>"}}. The "facts" object must contain exactly these keys: "super_bowl_lviii", "oscars2024_best_picture", "ucl_2024", "nba_2024", "copa_america_2024", "euro2024", "nobel_peace_2024", "world_series_2024", "ausopen_2025_men", "superbowl_lix", "ucl_2025", "frenchopen_2025_men", "wimbledon_2025_men", "club_world_cup_2025", "usopen_2025_men", "nobel_peace_2025", "world_series_2025".
- "super_bowl_lviii": Which NFL team won Super Bowl LVIII (played 2024-02-11)?
- "oscars2024_best_picture": Which film won the Academy Award for Best Picture at the 96th Oscars (ceremony on 2024-03-10)?
- "ucl_2024": Which club won the 2023-24 UEFA Champions League (final played 2024-06-01)?
- "nba_2024": Which team won the 2024 NBA Finals (concluded 2024-06-17)?
- "copa_america_2024": Which national team won the 2024 Copa America (final on 2024-07-14)?
- "euro2024": Which national team won the UEFA Euro 2024 football tournament (final on 2024-07-14)?
- "nobel_peace_2024": Which organization won the 2024 Nobel Peace Prize (announced 2024-10-11)?
- "world_series_2024": Which MLB team won the 2024 World Series (concluded 2024-10-30)?
- "ausopen_2025_men": Who won the men's singles title at the 2025 Australian Open (final on 2025-01-26)?
- "superbowl_lix": Which NFL team won Super Bowl LIX (played 2025-02-09)?
- "ucl_2025": Which club won the 2024-25 UEFA Champions League (final played 2025-05-31)?
- "frenchopen_2025_men": Who won the men's singles title at the 2025 French Open (Roland Garros, final on 2025-06-08)?
- "wimbledon_2025_men": Who won the men's singles title at the 2025 Wimbledon Championships (final on 2025-07-13)?
- "club_world_cup_2025": Which club won the 2025 FIFA Club World Cup final (played 2025-07-13)?
- "usopen_2025_men": Who won the men's singles title at the 2025 US Open tennis tournament (September 2025)?
- "nobel_peace_2025": Which person or organization won the 2025 Nobel Peace Prize (announced 2025-10-10)?
- "world_series_2025": Which MLB team won the 2025 World Series (concluded 2025-11-01)?

HTTP 头(TTFB) 18,730 ms 1,042 ms 891 ms 966 ms 971 ms

总耗时 20,114 ms 6,311 ms 4,458 ms 5,204 ms 13,674 ms

输入 1,252 8,126 7,567 8,126 1,508

输出 1,950 182 170 182 2,089

缓存读取 7,296 0 0 0 7,040

输出速度 burst 34.5 tok/s 47.7 tok/s 42.9 tok/s 164.4 tok/s

响应

{"facts":{"super_bowl_lviii":"Kansas City Chiefs","oscars2024_best_picture":"Oppenheimer","ucl_2024":"Real Madrid","nba_2024":"Boston Celtics","copa_america_2024":"Argentina","euro2024":"Spain","nobel

{"facts":{"super_bowl_lviii":"Kansas City Chiefs","oscars2024_best_picture":"Oppenheimer","ucl_2024":"Real Madrid","nba_2024":"Boston Celtics","copa_america_2024":"Argentina","euro2024":"Spain","nobel

{"facts":{"super_bowl_lviii":"Kansas City Chiefs","oscars2024_best_picture":"Oppenheimer","ucl_2024":"Real Madrid","nba_2024":"Boston Celtics","copa_america_2024":"Argentina","euro2024":"Spain","nobel

{"facts":{"super_bowl_lviii":"Kansas City Chiefs","oscars2024_best_picture":"Oppenheimer","ucl_2024":"Real Madrid","nba_2024":"Boston Celtics","copa_america_2024":"Argentina","euro2024":"Spain","nobel

{"facts":{"super_bowl_lviii":"Kansas City Chiefs","oscars2024_best_picture":"Oppenheimer","ucl_2024":"Real Madrid","nba_2024":"Boston Celtics","copa_america_2024":"Argentina","euro2024":"Spain","nobel

总计

总输入(含缓存) 49,583 49,181 36,596 49,180 40,244

总输出 2,097 354 307 322 2,145

总缓存创建 0 0 0 0 0

总缓存读取 38,016 38,912 22,656 29,184 33,536

输出速度 29.3 tok/s 34.2 tok/s 44.7 tok/s 41.5 tok/s 164.4 tok/s

总执行时间 1m 33.2s 1m 18.2s 1m 12.2s 1m 9.1s 1m 27.5s

总墙钟时间

9m 48.8s

含步骤间等待

19m 22.7s

含步骤间等待

9m 20.2s

含步骤间等待

21m 57.2s

含步骤间等待

9m 17.3s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.1398

OpenAI: GPT-5.5

$0.0814

OpenAI: GPT-5.5

$0.0902

OpenAI: GPT-5.5

$0.1242

OpenAI: GPT-5.5

$0.1147

OpenAI: GPT-5.5

通道指纹

来源步骤 knowledge_recall knowledge_recall knowledge_recall knowledge_recall knowledge_recall

平台 unknown (0%) unknown (0%) unknown (0%) unknown (0%) unknown (0%)

上游 - - - - -

CDN cloudflare cloudflare - - cloudflare

ID 格式 - - - - -

展开详细指纹（响应头 / body 特征 / 完整信号列表）

通道	response_headers_notable	response_body_traits	signals	request_id_chain
O-Plus	cf-ray: a14ec7f60e48f5b8-AMS server: cloudflare	—	hdr:server:cloudflarehdr:cf-ray	89a6b3da-c20b-4059-81df-8bad38640bd4 a14ec7f60e48f5b8-AMS
cx	cf-ray: a14ed6441e0fcbe0-FRA server: cloudflare	—	hdr:server:cloudflarehdr:cf-ray	9ceb1027-f7ab-4ac4-b11b-2e1a0ab5e278 a14ed6441e0fcbe0-FRA
O-Team	server: openresty	—	—	299cb1e8-3af2-4f94-8857-1838615f9538
官方基准	server: nginx/1.22.1	—	—	—
u-cx	via: 1.1 Caddy cf-ray: a14ec7492b57fd4f-NRT server: cloudflare	via_chain: ["1.1 Caddy"]	hdr:server:cloudflarehdr:cf-rayhdr:via	a14ec7492b57fd4f-NRT

协议指纹评分（机器自动）

总分 / 100 97 100 83 99 87

基线状态已对照基线已对照基线已对照基线已对照基线已对照基线

缓存命中比 10.0 /10 ×20 10.0 /10 ×20 10.0 /10 ×20 9.3 /10 ×20 10.0 /10 ×20

延迟基线 8.3 /10 ×15 10.0 /10 ×15 8.0 /10 ×15 10.0 /10 ×15 7.5 /10 ×15

模型匹配 10.0 /10 ×14 10.0 /10 ×14 10.0 /10 ×14 10.0 /10 ×14 10.0 /10 ×14

系统提示纯净 10.0 /10 ×8 10.0 /10 ×8 0.0 /10 ×8 10.0 /10 ×8 0.0 /10 ×8

身份(结构) 10.0 /10 ×7 10.0 /10 ×7 10.0 /10 ×7 10.0 /10 ×7 10.0 /10 ×7

身份(自由) 10.0 /10 ×7 10.0 /10 ×7 10.0 /10 ×7 10.0 /10 ×7 10.0 /10 ×7

知识截止 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5

流式投递未计入未计入未计入未计入未计入

另有 9 个维度不适用于本服务，已排除评分（不计入总分）

综合结论

与基线相似度 97 v3.27.0 100 v3.27.0 83 v3.27.0 基准（参考） 87 v3.27.0

雷达图例

共 8 个维度（顺时针，从顶部 12 点起）— 点击展开对照

1 模型匹配 model_match
2 知识截止 cutoff_match
3 自由身份 identity_free_clean
4 系统提示纯净 system_prompt_clean
5 缓存命中比 cache_hit_ratio_match
6 知识召回 knowledge_recall_match
7 延迟基线 latency_baseline_match
8 结构化身份 identity_structured_match

每根轴长度 = 该维度 0-10 分（越长越接近基线）。蓝色虚线圆 = 满分基准。具体权重见方法论页。

维度雷达

平均延迟 6997 ms 4383 ms 3010 ms 3080 ms 6938 ms

输出速度 29.3 tok/s 34.2 tok/s 44.7 tok/s 41.5 tok/s 164.4 tok/s

完成步骤 5/5 成功 5/5 成功 5/5 成功 5/5 成功 5/5 成功

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.1398

$0.0814

$0.0902

$0.1242

$0.1147

缓存读取占比 ⓘ 77% 79% 62% 59% 83%

揭晓

通道

服务商 AheiAPI ↗

通道O-Plus

服务商 YesCode ↗

通道cx

服务商 CodexApi ↗

通道O-Team

官方基准

服务商 right ↗

通道u-cx

返回列表