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

全部 ≥ 70 与基线相似度

已揭晓揭晓进度

步骤 / 指标

JikeAI

RayinCode

O-Pro

官方基准

iKunAPI

O-Pro20X

SAIAi

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 16:06:57 ~ 16:07:17 (18.1s) 16:06:53 ~ 16:07:13 (16.7s) 16:07:01 ~ 16:07:16 (12.5s) 16:06:55 ~ 16:07:10 (11.8s) 16:06:59 ~ 16:07:14 (12.8s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 2,342 ms 4,438 ms 849 ms 1,249 ms 1,446 ms

总耗时 5,410 ms 5,661 ms 1,595 ms 1,449 ms 2,153 ms

输入 1,579 1,548 1,548 1,548 1,548

输出 5 5 5 5 5

缓存读取 8,576 8,576 8,576 8,576 8,576

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 16:07:34 ~ 16:07:54 (19.1s) 16:07:31 ~ 16:07:46 (14.9s) 16:07:57 ~ 16:08:12 (14.7s) 16:07:40 ~ 16:07:55 (12.6s) 16:07:43 ~ 16:08:03 (16.8s)

提示词

查看本步 Prompt

输出 3 行 metadata 描述你自己（当前 API 模型）。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。禁止任何额外字符：
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) 4,088 ms 1,312 ms 860 ms 1,401 ms 959 ms

总耗时 8,370 ms 3,749 ms 4,058 ms 2,153 ms 6,105 ms

输入 7,859 148 1,684 1,684 1,684

输出 245 102 23 20 131

缓存读取 2,432 10,112 8,576 8,576 8,576

输出速度 57.2 tok/s 41.9 tok/s - - 25.5 tok/s

响应

vendor: openai
brand: gpt
model: gpt-5

vendor: openai
brand: codex
model: gpt-5

vendor: openai
brand: codex
model: gpt-5-codex

vendor: openai
brand: codex
model: gpt-5

vendor: openai
brand: codex
model: gpt-5-codex

步骤 3: cutoff text

时间 16:08:12 ~ 16:08:37 (22.0s) 16:08:10 ~ 16:08:25 (12.8s) 16:08:39 ~ 16:08:54 (12.5s) 16:08:16 ~ 16:08:31 (12.3s) 16:08:33 ~ 16:08:48 (13.2s)

提示词

查看本步 Prompt

你掌握的世界知识截止于何时？YYYY-MM 格式作答，不解释。

HTTP 头(TTFB) 2,019 ms 1,439 ms 954 ms 1,660 ms 1,006 ms

总耗时 11,023 ms 2,255 ms 1,871 ms 1,892 ms 2,491 ms

输入 737 1,730 1,733 1,730 1,731

输出 231 8 8 8 8

缓存读取 9,600 8,576 8,576 8,576 8,576

输出速度 25.7 tok/s - - - -

响应

2024-06

2024-06

2024-06

2024-06

2024-06

步骤 4: identity_free text

时间 16:08:58 ~ 16:09:13 (13.7s) 16:09:00 ~ 16:10:25 (81.9s) 16:09:28 ~ 16:09:43 (13.5s) 16:09:14 ~ 16:09:29 (12.5s) 16:09:22 ~ 16:09:42 (16.2s)

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) 1,831 ms - 955 ms 1,543 ms 2,506 ms

总耗时 2,836 ms - 2,807 ms 2,173 ms 5,525 ms

输入 257 - 1,765 225 224

输出 26 - 26 23 31

缓存读取 10,112 - 8,576 10,112 10,112

输出速度 - - - - 10.3 tok/s

响应

我是 OpenAI 开发的 AI 助手，提供对话、写作和编程支持。

我是 OpenAI 开发的 Codex 编程助手，可协助阅读、修改、测试代码并解决软件工程问题。

我是 OpenAI 开发的 Codex，一款面向编程协作的 AI 助手。

我叫 Codex，属于 OpenAI 的 Codex/API 产品，由 OpenAI 开发。

我是 OpenAI 开发的 Codex，基于 GPT-5，用于协助编程、分析与自动化任务。

步骤 5: knowledge_recall text

时间 16:09:41 ~ 16:10:06 (20.0s) 16:11:04 ~ 16:11:24 (17.7s) 16:10:18 ~ 16:10:38 (15.5s) 16:09:58 ~ 16:10:23 (23.7s) 16:10:02 ~ 16:10:22 (16.8s)

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) 4,410 ms 2,304 ms 941 ms 4,589 ms 1,986 ms

总耗时 7,646 ms 6,017 ms 4,882 ms 12,317 ms 6,067 ms

输入 477 8,126 8,126 8,126 8,126

输出 185 182 179 182 179

缓存读取 7,680 0 0 0 0

输出速度 57.2 tok/s 49.0 tok/s 45.4 tok/s 23.6 tok/s 43.9 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) 49,309 38,816 49,160 49,153 49,153

总输出 692 297 241 238 354

总缓存创建 0 0 0 0 0

总缓存读取 38,400 27,264 34,304 35,840 35,840

输出速度 40.0 tok/s 46.2 tok/s 45.4 tok/s 23.6 tok/s 27.8 tok/s

总执行时间 1m 32.8s 2m 24.0s 1m 8.7s 1m 12.9s 1m 15.7s

总墙钟时间

3m 8.7s

含步骤间等待

4m 30.8s

含步骤间等待

3m 36.9s

含步骤间等待

3m 27.9s

含步骤间等待

3m 22.9s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.0945

OpenAI: GPT-5.5

$0.0803

OpenAI: GPT-5.5

$0.0987

OpenAI: GPT-5.5

$0.0916

OpenAI: GPT-5.5

$0.0951

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%) unknown (0%) one-api (100%) 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
cx	cf-ray: a14c2b803c5ac0b5-FRA server: cloudflare set-cookie: p_uv_id=70a21be30c63c87f8762d8dbdf0a2648; path=/; expires=Fri, 03-Jul-26 00:00:00 GMT; HttpOnly x-oneapi-request-id: 202607020809442036428028268d9d6M52iWjZz	—	hdr:x-oneapi-request-idhdr:server:cloudflarehdr:cf-ray	a14c2b803c5ac0b5-FRA
O-Pro	cf-ray: a14c2d83d9253fd6-SIN server: cloudflare	—	hdr:server:cloudflarehdr:cf-ray	18707452-056e-4615-98fb-32e15c6ea632 a14c2d83d9253fd6-SIN
官方基准	server: nginx/1.22.1	—	—	—
O-Pro20X	server: nginx x-new-api-version: v1.0.0-rc.12 x-oneapi-request-id: 202607020809598687656918268d9d6OCYGwVL1	—	hdr:x-new-api-versionhdr:x-oneapi-request-id	—
O-Pro	server: nginx/1.22.1	—	—	—

协议指纹评分（机器自动）

总分 / 100 89 100 99 99 98

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 3.4 /10 ×15 9.9 /10 ×15 10.0 /10 ×15 9.6 /10 ×15 9.7 /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 10.0 /10 ×8 10.0 /10 ×8 10.0 /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 89 v3.26.0 100 v3.26.0 基准（参考） 99 v3.26.0 98 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 7057 ms 3536 ms 3043 ms 3997 ms 4468 ms

输出速度 40.0 tok/s 46.2 tok/s 45.4 tok/s 23.6 tok/s 27.8 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 ⓘ

$0.0945

$0.0803

$0.0987

$0.0916

$0.0951

缓存读取占比 ⓘ 78% 70% 70% 73% 73%

揭晓

通道

服务商 JikeAI ↗

通道cx

服务商 RayinCode ↗

通道O-Pro

官方基准

服务商 iKunAPI ↗

通道O-Pro20X

服务商 SAIAi ↗

通道O-Pro

返回列表