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

sucui

O-Pro

官方基准

KudeX

M-Plus

YunDou

O-Pro

AiTokensFlux

O-Pro

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 21:18:42 ~ 21:19:07 (21.5s) 21:18:46 ~ 21:19:01 (12.2s) 21:18:44 ~ 21:18:59 (14.0s) 21:18:38 ~ 21:18:53 (12.6s) 21:18:40 ~ 21:18:55 (12.0s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 9,137 ms 988 ms 2,453 ms 930 ms 908 ms

总耗时 10,590 ms 1,559 ms 2,689 ms 1,905 ms 1,559 ms

输入 1,548 1,548 2,322 1,548 7,910

输出 5 5 8 5 5

缓存读取 8,576 8,576 12,864 8,576 2,432

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 21:19:31 ~ 21:19:51 (16.0s) 21:19:38 ~ 21:19:53 (14.5s) 21:19:19 ~ 21:19:34 (13.3s) 21:19:15 ~ 21:19:30 (13.9s) 21:19:11 ~ 21:19:26 (14.4s)

提示词

查看本步 Prompt

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。严格按以下 3 行格式输出，全部小写英文，不要任何其他文字、解释、标点：
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) 1,157 ms 1,014 ms 2,336 ms 914 ms 1,077 ms

总耗时 5,276 ms 3,845 ms 2,823 ms 2,051 ms 3,842 ms

输入 151 1,687 226 151 369

输出 131 123 30 20 107

缓存读取 10,112 8,576 15,168 10,112 10,112

输出速度 31.8 tok/s 43.4 tok/s 61.6 tok/s - 38.7 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 21:20:08 ~ 21:20:28 (18.1s) 21:20:32 ~ 21:20:47 (12.3s) 21:20:02 ~ 21:20:17 (14.3s) 21:19:58 ~ 21:20:13 (12.4s) 21:20:05 ~ 21:20:20 (12.5s)

提示词

查看本步 Prompt

Knowledge cutoff date? Reply with YYYY-MM only.

HTTP 头(TTFB) 1,482 ms 938 ms 3,435 ms 886 ms 1,016 ms

总耗时 7,393 ms 1,653 ms 3,781 ms 1,658 ms 1,993 ms

输入 1,723 197 2,596 187 1,941

输出 8 8 12 8 8

缓存读取 8,576 10,112 12,864 10,112 8,576

响应

2024-06

2024-06

2024-06

2024-06

2024-06

步骤 4: identity_free text

时间 21:21:13 ~ 21:21:28 (14.4s) 21:21:15 ~ 21:21:30 (13.1s) 21:20:59 ~ 21:21:14 (14.1s) 21:20:34 ~ 21:20:49 (12.6s) 21:20:52 ~ 21:21:07 (13.4s)

提示词

查看本步 Prompt

请用中文介绍你自己是什么产品或助手，由谁开发，50 字以内。

HTTP 头(TTFB) 1,154 ms 980 ms 3,057 ms - 1,062 ms

总耗时 2,963 ms 2,373 ms 3,615 ms - 2,772 ms

输入 1,755 229 339 - 437

输出 35 23 33 - 33

缓存读取 8,576 10,112 15,168 - 10,112

输出速度 19.3 tok/s - 59.1 tok/s - 19.3 tok/s

响应

我是 OpenAI 开发的 Codex 编程助手，基于 GPT-5，帮助阅读、修改、测试代码并解答技术问题。

我是 Codex，基于 GPT-5 的编程助手，由 OpenAI 打造。

我叫 Codex，属于 OpenAI 的编码助手产品，由 OpenAI 开发。

我是 OpenAI 开发的 Codex，基于 GPT-5 的编码助手，可协助阅读、修改、测试和解释代码。

我是 Codex，基于 GPT-5 的编程助手，由 OpenAI 打造，用于协助代码理解、修改与验证。

步骤 5: knowledge_recall text

时间 21:21:53 ~ 21:22:13 (17.1s) 21:22:09 ~ 21:22:29 (16.3s) 21:21:30 ~ 21:21:50 (16.9s) 21:21:23 ~ 21:21:43 (15.4s) 21:21:28 ~ 21:21:58 (29.8s)

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) 1,273 ms 967 ms 2,943 ms 885 ms 6,397 ms

总耗时 6,059 ms 5,523 ms 6,369 ms 4,726 ms 19,292 ms

输入 8,126 446 12,189 8,126 8,422

输出 179 182 273 182 680

缓存读取 0 7,680 0 0 0

输出速度 37.4 tok/s 39.9 tok/s 79.7 tok/s 47.4 tok/s 52.7 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":"UNKNOWN","copa_america_2024":"UNKNOWN","euro2024":"UNKNOWN","nobel_peace_

总计

总输入(含缓存) 49,143 49,163 73,736 38,812 50,311

总输出 358 341 356 215 833

总缓存创建 0 0 0 0 0

总缓存读取 35,840 45,056 56,064 28,800 31,232

输出速度 32.2 tok/s 41.3 tok/s 75.2 tok/s 47.4 tok/s 47.2 tok/s

总执行时间 1m 27.1s 1m 8.4s 1m 12.6s 1m 6.9s 1m 22.2s

总墙钟时间

3m 30.9s

含步骤间等待

3m 42.9s

含步骤间等待

3m 5.9s

含步骤间等待

3m 5.7s

含步骤间等待

3m 17.9s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.0952

OpenAI: GPT-5.5

$0.0533

OpenAI: GPT-5.5

$0.1271

OpenAI: GPT-5.5

$0.0709

OpenAI: GPT-5.5

$0.1360

OpenAI: GPT-5.5

通道指纹

来源步骤 knowledge_recall knowledge_recall knowledge_recall knowledge_recall knowledge_recall

平台 one-api (100%) unknown (0%) one-api (100%) unknown (0%) unknown (0%)

上游 - - - - -

CDN 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-Pro	via: 1.1 Caddy cf-ray: a14df4c91ce4dc9d-FRA server: cloudflare x-new-api-version: v0.12.10 x-oneapi-request-id: 202607021321545668494318268d9d6gi3aFhgo	via_chain: ["1.1 Caddy"]	hdr:x-new-api-versionhdr:x-oneapi-request-idhdr:server:cloudflarehdr:cf-rayhdr:via	a14df4c91ce4dc9d-FRA
官方基准	server: nginx/1.22.1	—	—	—
M-Plus	cf-ray: a14df4364a6cc8eb-NRT server: cloudflare x-oneapi-request-id: 202607021321308228150228268d9d6sExVI1pM	—	hdr:x-oneapi-request-idhdr:server:cloudflarehdr:cf-ray	a14df4364a6cc8eb-NRT
O-Pro	server: nginx	—	—	6a1fbf7a-01aa-4852-bc3c-9726ea2b3452
O-Pro	server: nginx	—	—	—

协议指纹评分（机器自动）

总分 / 100 96 100 83 100 86

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 10.0 /10 ×20 10.0 /10 ×20

延迟基线 7.4 /10 ×15 10.0 /10 ×15 5.5 /10 ×15 10.0 /10 ×15 10.0 /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 5.0 /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 96 v3.26.0 基准（参考） 83 v3.26.0 100 v3.26.0 86 v3.26.0

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 6456 ms 2991 ms 3855 ms 2068 ms 5892 ms

输出速度 32.2 tok/s 41.3 tok/s 75.2 tok/s 47.4 tok/s 47.2 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 ⓘ

$0.0952

$0.0533

$0.1271

$0.0709

$0.1360

缓存读取占比 ⓘ 73% 92% 76% 74% 62%

揭晓

通道

服务商 sucui ↗

通道O-Pro

官方基准

服务商 KudeX ↗

通道M-Plus

服务商 YunDou ↗

通道O-Pro

服务商 AiTokensFlux ↗

通道O-Pro

返回列表