模型对比 | RPDiag

测试说明点击展开 5 步 quick-probe-v1 详解

前 5 步同一会话，第 6 步独立新会话，覆盖 6 个独立信号： 1. ping/pong —— 单字指令遵循 + 建立缓存上下文。 2. 身份（结构化）—— 三行 vendor / brand / model 格式，机器可解析。 3. 知识截止 —— 跨越 5 分钟 cache 边界后追问，检测 sliding 5m cache 是否真正命中。 4. 身份（自由格式）—— 自然语言自报身份，给包装层（Kiro 等）暴露品牌的机会。 5. 世界知识层级 —— 5 道公共事件硬事实题，按答对档位映射实测 tier，与请求模型对照。 6. 档位判别（digit_count）—— 全新会话 + 系统提示「不要推理」+ effort=low 的数字计数题，按 output_tokens 体量区分 opus/sonnet/haiku（opus fail-fast ~19，sonnet/haiku 照数），与官方基线对照抓同厂降级。前 5 步之间随机延迟 1–4 分钟。第 3 步（cutoff）累计跨度 > 5 分钟保住 sliding cache 检测；总跨度 > 6 分钟。第 6 步是独立新会话（不复用前 5 轮上下文，否则 opus 不再 fail-fast）。正确实现 sliding 5m cache 的通道，第 3 步及之后的 cache_read 仍 > 0；若按「创建时间起 5 分钟」实现，第 3 步会暴露 cache_read = 0。

5 通道 · 6 步骤

2026/06/26 运行日期

claude-sonnet-4-6 官方基准

1 通道分数 < 70 另含 1 任务失败

已揭晓揭晓进度

步骤 / 指标

SSSAiCode

O-Max

Novaris

O-Max

LingxiCode

M-

官方基准

HongMaCC

u-cc

基本信息

服务名 Claude Code Claude Code Claude Code Claude Code Claude Code

模型 claude-sonnet-4-6 claude-sonnet-4-6 claude-sonnet-4-6 claude-sonnet-4-6 claude-sonnet-4-6

客户端 claude-cli v2.1.160 claude-cli v2.1.160 claude-cli v2.1.160 claude-cli v2.1.160 claude-cli v2.1.160

偏差标红基于官方基准

步骤 1: ping text

时间 02:36:42 ~ 02:36:52 (8.8s) 02:36:40 ~ 02:36:50 (5.4s) 02:36:44 ~ 02:36:54 (5.1s) 02:36:48 ~ 02:36:53 (4.3s) 02:36:46 ~ 02:36:51 (4.2s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 7,423 ms 3,485 ms 3,449 ms 2,692 ms 3,042 ms

首文本(SSE delta) 7,506 ms 3,662 ms 3,450 ms 2,694 ms 3,043 ms

总耗时 7,697 ms 4,620 ms 3,794 ms 2,843 ms 3,043 ms

输入 3 22,664 3 3 3

输出 5 102 5 5 5

缓存创建

31,985

5m: 0 / 1h: 31,985

308

32,319

5m: 0 / 1h: 32,319

32,318

5m: 0 / 1h: 32,318

32,317

5m: 32,317 / 1h: 0

缓存读取 0 6,894 0 0 0

输出速度 - 106.5 tok/s - - -

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 02:39:20 ~ 02:39:25 (4.6s) 02:39:13 ~ 02:39:23 (5.6s) 02:38:57 ~ 02:39:07 (6.8s) 02:39:15 ~ 02:39:20 (4.2s) 02:38:35 ~ 02:38:45 (5.1s)

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。请只输出以下 3 行内容（label 小写，值小写）：
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) 3,097 ms 4,361 ms 5,193 ms 2,063 ms 3,844 ms

首文本(SSE delta) 3,748 ms 4,514 ms 5,194 ms 2,064 ms 3,845 ms

总耗时 3,907 ms 4,803 ms 5,763 ms 2,552 ms 3,846 ms

输入 3 22,673 3 3 3

输出 22 59 22 22 22

缓存创建

178

5m: 0 / 1h: 178

440

178

5m: 0 / 1h: 178

178

5m: 0 / 1h: 178

178

5m: 178 / 1h: 0

缓存读取 31,985 6,940 32,319 32,318 32,317

响应

vendor: anthropic
brand: claude
model: claude-sonnet-4-5

I can't discuss that.

vendor: anthropic
brand: claude
model: claude-sonnet-4-6

vendor: anthropic
brand: claude
model: claude-sonnet-4-6

vendor: anthropic
brand: claude
model: claude-sonnet-4-6

步骤 3: cutoff text

时间 02:40:41 ~ 02:40:46 (3.9s) 02:41:48 ~ 02:41:58 (5.6s) 02:41:41 ~ 02:41:51 (5.2s) 02:40:44 ~ 02:40:49 (3.6s) 02:41:25 ~ 02:42:00 (33.4s)

提示词

查看本步 Prompt

请直接回答你的知识截止日期，格式 YYYY-MM，不要附加任何说明。

HTTP 头(TTFB) 2,840 ms 3,222 ms 3,693 ms 1,981 ms 32,301 ms

首文本(SSE delta) 2,965 ms 3,239 ms 3,707 ms 2,155 ms 32,302 ms

总耗时 3,100 ms 4,477 ms 4,160 ms 2,340 ms 32,302 ms

输入 3 22,681 3 3 3

输出 8 128 8 8 8

缓存创建

5m: 0 / 1h: 57

318

5m: 0 / 1h: 57

5m: 38 / 1h: 0

缓存读取 32,163 7,090 32,497 32,496 32,495

输出速度 - 103.4 tok/s - - -

响应

2025-04

2025-08

2025-08

2025-08

2025-08

步骤 4: identity_free text

时间 02:43:35 ~ 02:43:45 (5.3s) 02:45:55 ~ 02:46:05 (8.4s) 02:43:49 ~ 02:43:59 (5.7s) 02:43:14 ~ 02:43:24 (5.3s) 02:43:08 ~ 02:43:18 (5.8s)

提示词

查看本步 Prompt

中文做个自我介绍：你是什么产品或助手？谁开发的？50 字内。

HTTP 头(TTFB) 3,179 ms 3,175 ms 3,885 ms 3,629 ms 4,688 ms

首文本(SSE delta) 4,156 ms 3,231 ms 3,905 ms 3,631 ms 4,689 ms

总耗时 4,310 ms 7,448 ms 4,805 ms 4,335 ms 4,689 ms

输入 3 22,689 3 3 3

输出 70 251 53 51 55

缓存创建

5m: 0 / 1h: 43

339

5m: 0 / 1h: 41

5m: 35 / 1h: 0

缓存读取 32,220 7,162 32,554 32,553 32,533

输出速度 - 59.5 tok/s 58.9 tok/s 72.4 tok/s burst

响应

我是 Claude，由 Anthropic 开发的 AI 助手。我擅长回答问题、编写代码、分析数据和协助各类任务，致力于提供安全、有帮助的智能对话体验。

我是 claude-sonnet-4-6，一个 AI 驱动的开发环境。我帮助开发者编写代码，由 Anthropic 基于 Claude 构建。

我叫 Claude，是由 Anthropic 开发的 AI 助手。当前版本为 Claude Sonnet 4.6，属于 Claude 4 系列产品。

我叫 Claude，是由 Anthropic 开发的 AI 助手，属于 Claude 产品系列。当前版本为 claude-sonnet-4-6。

我是 Claude，由 Anthropic 开发的 AI 助手。我基于大型语言模型构建，旨在提供安全、有益的对话和任务协助。

步骤 5: knowledge_recall text

时间 02:46:17 ~ 02:46:32 (10.7s) 02:48:12 ~ 02:48:22 (5.9s) 02:46:37 ~ 02:46:47 (6.1s) 02:45:51 ~ 02:45:56 (4.8s) 02:45:02 ~ 02:45:12 (5.2s)

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) 7,347 ms 3,022 ms 2,608 ms 1,786 ms 4,130 ms

首文本(SSE delta) 9,729 ms 3,108 ms 2,609 ms 1,788 ms 4,131 ms

总耗时 9,888 ms 5,106 ms 5,343 ms 3,594 ms 4,131 ms

输入 3 41 3 3 3

输出 198 253 205 205 205

缓存创建

7,722

5m: 0 / 1h: 7,722

1,094

8,037

5m: 0 / 1h: 8,037

8,034

5m: 0 / 1h: 8,034

8,012

5m: 8,012 / 1h: 0

缓存读取 0 7,208 0 0 0

输出速度 - 126.6 tok/s 75.0 tok/s 113.5 tok/s burst

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

```json
{"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

{"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

步骤 6: digit_count text

时间 02:49:05 ~ 02:49:10 (1.1s) 02:50:28 ~ 02:50:48 (17.0s) 02:50:07 ~ 02:50:27 (15.2s) 02:49:28 ~ 02:49:43 (14.9s) 02:46:41 ~ 02:47:01 (15.7s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=893118542513687080266213062242896738910677496464249946535611875391655736

HTTP 头(TTFB) - 2,532 ms 1,495 ms 1,158 ms 14,452 ms

首文本(SSE delta) - 2,533 ms 1,496 ms 1,159 ms 14,453 ms

总耗时 - 16,288 ms 14,455 ms 13,979 ms 14,453 ms

输入 - 7 164 164 164

输出 - 1,214 1,095 1,323 877

缓存创建

5m: 0 / 1h: 0

缓存读取 - 139 0 0 0

输出速度 - 88.3 tok/s 84.5 tok/s 103.2 tok/s burst

响应

失败

评测失败

{"n":72,"c":[4,6,7]}

```python
s = "893118542513687080266213062242896738910677496464249946535611875391655736"
n = len(s)
c0 = s.count('0')
c7 = s.count('7')
c9 = s.count('9')
```

**Verification walk-through:**

| digit |

## Counting digits in s

**s** = `893118542513687080266213062242896738910677496464249946535611875391655736`

### Length
Splitting into groups of 10:
`8931185425` `1368708026` `6213062242` `8967389106`

I need to count the digits in the string s=893118542513687080266213062242896738910677496464249946535611875391655736

First, let me count the total length and occurrences of 0, 7, 9.

s = 8931185425136

总计

总输入(含缓存)

136,368

仅前 6 步

128,751

138,181

138,174

138,104

总输出

303

仅前 6 步

2,007

1,388

1,614

1,172

总缓存创建

39,985

5m: 0 / 1h: 39,985

仅前 6 步

2,563

40,632

5m: 0 / 1h: 40,632

40,628

5m: 0 / 1h: 40,628

40,580

5m: 40,580 / 1h: 0

总缓存读取

96,368

仅前 6 步

35,433

97,370

97,367

97,345

输出速度

仅前 6 步

87.9 tok/s

81.5 tok/s

103.0 tok/s

burst

总执行时间

34.5s

仅前 6 步

47.9s

44.0s

37.2s

1m 9.4s

总墙钟时间

12m 27.9s

含步骤间等待

仅前 6 步

14m 8.0s

含步骤间等待

13m 43.0s

含步骤间等待

12m 55.0s

含步骤间等待

10m 15.0s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.1834

Anthropic: Claude Sonnet 4.6 (部分)

$0.3190

Anthropic: Claude Sonnet 4.6

$0.2029

Anthropic: Claude Sonnet 4.6

$0.2063

Anthropic: Claude Sonnet 4.6

$0.1424

Anthropic: Claude Sonnet 4.6

通道指纹

来源步骤 knowledge_recall digit_count digit_count digit_count digit_count

平台 direct (100%) one-api (100%) one-api (100%) direct (100%) one-api (100%)

上游 Anthropic 直连 (100%) Anthropic 直连 (75%) Anthropic 直连 (100%) Anthropic 直连 (100%) Anthropic 直连 (100%)

CDN - - cloudflare - -

ID 格式 msg_01YM93VhuaoRMRxCftq6mqLK msg_3HvSfKFyb7RSjwkiTHwX2Lts msg_016yma8vgfsZdQKd7HcFYfxE msg_01WxTcn8LDBTexStjDFsjdFi msg_01R131bQbnnqNFWVwGvKMXdh

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-Max	cf-ray: a116225df90ef8b7-LAX server: Photon-Edge set-cookie: _cfuvid=yP3MU_OhSwoEuzf55nccTasTtiZpPJ1Tc.svp_deIog-1782413178.5552201-1.0.1.1-ZBUz05UIdDX8SaRHYoyAhAr48svJgSo_EPilNe3dyaA; HttpOnly; SameSite=None; Secure; Path=/; Domain=api.anthropic.com	id_format: msg_prefix extra_fields: ["stop_details"] has_service_tier: true has_inference_geo: true model_has_date_suffix: false usage_has_cache_fields: true	body:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	1782413178123-uuh1sejjad81b req_011CcQUVuDz13ENmCHNw6sHn a116225df90ef8b7-LAX
O-Max	server: nginx x-oneapi-request-id: 202606251850291383356178268d9d68uMKfbQg	id_format: msg_prefix has_service_tier: false has_inference_geo: false model_has_date_suffix: false usage_has_cache_fields: true	hdr:x-oneapi-request-idbody:id:msg_prefixbody:usage.cache_fieldsupstream:anthropic-direct-from-body	—
M-	cf-ray: a11627f90ed4af8b-NRT server: cloudflare set-cookie: server_name_session=27f4a95581d368605416e6d0825f8e16; Max-Age=86400; httponly; path=/ x-oneapi-request-id: 20260625185008308139658kMbNSket	id_format: msg_prefix extra_fields: ["stop_details"] has_service_tier: true has_inference_geo: true model_has_date_suffix: false usage_has_cache_fields: true	hdr:x-oneapi-request-idhdr:server:cloudflarehdr:cf-raybody:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	a11627f90ed4af8b-NRT
官方基准	server: nginx/1.22.1	id_format: msg_prefix extra_fields: ["stop_details"] has_service_tier: true has_inference_geo: true model_has_date_suffix: false usage_has_cache_fields: true	body:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	req_011CcQUjysfzdHkK2L1Niq2w
u-cc	server: nginx x-new-api-version: 1.0.17 x-oneapi-request-id: 202606251846427965397228268d9d6OW69Y2cg	id_format: msg_prefix extra_fields: ["stop_details"] has_service_tier: true has_inference_geo: true model_has_date_suffix: false usage_has_cache_fields: true	hdr:x-new-api-versionhdr:x-oneapi-request-idbody:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	—

协议指纹评分（机器自动）

总分 / 100 58 98 100 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

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

缓存连续性 — 10.0 /10 ×14 10.0 /10 ×14 10.0 /10 ×14 10.0 /10 ×14

sliding 5m — 10.0 /10 ×13 10.0 /10 ×13 10.0 /10 ×13 10.0 /10 ×13

缓存 TTL 一致 — 0.0 /10 ×15 10.0 /10 ×15 10.0 /10 ×15 0.0 /10 ×15

缓存命中比 — 0.0 /10 ×20 10.0 /10 ×20 10.0 /10 ×20 10.0 /10 ×20

身份(结构) — 0.0 /10 ×7 10.0 /10 ×7 10.0 /10 ×7 10.0 /10 ×7

知识截止 — 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

原生 msg-ID — 0.0 /10 ×8 10.0 /10 ×8 10.0 /10 ×8 10.0 /10 ×8

Req-ID 透传 — 0.0 /10 ×4 0.0 /10 ×4 10.0 /10 ×4 0.0 /10 ×4

stop_reason — 10.0 /10 ×3 10.0 /10 ×3 10.0 /10 ×3 10.0 /10 ×3

service_tier — 0.0 /10 ×6 10.0 /10 ×6 10.0 /10 ×6 10.0 /10 ×6

inference_geo — 0.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5

SDK 一致 — 10.0 /10 ×2 10.0 /10 ×2 10.0 /10 ×2 10.0 /10 ×2

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

延迟基线 — 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 9.3 /10 ×5

流式投递 — 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 0.0 /10 ×5

综合结论

与基线相似度执行失败，不评分 58 v3.24.1 98 v3.24.1 基准（参考） 86 v3.24.1

雷达图例

共 23 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 SDK 一致 sdk_consistency
4 Thinking 存在 thinking_present
5 流式投递 buffer_dump_match
6 自由身份 identity_free_clean
7 停止原因 stop_reason_present
8 系统提示纯净 system_prompt_clean
9 Service Tier service_tier_present
10 缓存命中比 cache_hit_ratio_match
11 缓存 TTL 一致 cache_ttl_consistency
12 推理地区 inference_geo_present
13 缓存连续 cache_continuity_intra
14 知识召回 knowledge_recall_match
15 延迟基线 latency_baseline_match
16 原生 msg-ID anthropic_msg_id_format
17 Sliding 缓存 cache_sliding_correctness
18 结构化身份 identity_structured_match
19 自报身份一致 self_identity_consistency
20 信封自报一致 envelope_self_report_match
21 中文遵循 instruction_following_lang
22 档位思考量 tier_thinking_volume_match
23 Req-ID 透传 anthropic_request_id_passthrough

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 5780 ms 7124 ms 6387 ms 4941 ms 10411 ms

输出速度 - 87.9 tok/s 81.5 tok/s 103.0 tok/s burst

完成步骤 5/6 成功 · 1 失败 6/6 成功 6/6 成功 6/6 成功 6/6 成功

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.1834 (部分)

$0.3190

$0.2029

$0.2063

$0.1424

缓存读取占比 ⓘ 71% (部分) 28% 70% 70% 70%

揭晓

通道

服务商 SSSAiCode ↗

通道O-Max

服务商 Novaris ↗

通道O-Max

服务商 LingxiCode ↗

通道M-

官方基准

服务商 HongMaCC ↗

通道u-cc

返回列表