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

2026/06/29 运行日期

claude-sonnet-4-6 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

TopRouterCN

O-Max

ClaudeCN

O-Max

FoxCode

R-super

官方基准

ByteCatCode

O-Max

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 15:10:00 ~ 15:10:05 (4.5s) 15:10:02 ~ 15:10:12 (8.2s) 15:10:04 ~ 15:10:09 (3.6s) 15:10:06 ~ 15:10:16 (4.9s) 15:09:58 ~ 15:10:08 (5.1s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 3,517 ms 5,749 ms 2,159 ms 3,697 ms 3,545 ms

首文本(SSE delta) 3,518 ms 5,750 ms 2,159 ms 3,699 ms -

总耗时 3,628 ms 6,168 ms 2,735 ms 3,857 ms 3,794 ms

输入 1,467 3 57 3 3

输出 5 5 14 5 5

缓存创建

30,854

5m: 0 / 1h: 30,854

32,319

5m: 0 / 1h: 32,319

32,958

5m: 32,958 / 1h: 0

32,318

5m: 0 / 1h: 32,318

32,457

5m: 0 / 1h: 32,457

缓存读取 0 0 0 0 0

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 15:11:17 ~ 15:11:22 (3.9s) 15:11:43 ~ 15:11:53 (5.5s) 15:11:33 ~ 15:11:43 (5.3s) 15:12:09 ~ 15:12:14 (3.4s) 15:11:36 ~ 15:11:46 (6.0s)

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) 3,106 ms 3,882 ms 2,257 ms 2,026 ms 3,721 ms

首文本(SSE delta) 3,107 ms 3,883 ms 2,258 ms 2,027 ms -

总耗时 3,191 ms 4,279 ms 4,344 ms 2,421 ms 4,752 ms

输入 3 3 57 3 3

输出 22 22 85 22 22

缓存创建

1,666

5m: 0 / 1h: 1,666

202

5m: 0 / 1h: 202

533

5m: 533 / 1h: 0

202

5m: 0 / 1h: 202

202

5m: 0 / 1h: 202

缓存读取 30,854 32,319 32,958 32,318 32,457

输出速度 - - 40.7 tok/s - -

响应

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

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 15:13:54 ~ 15:14:04 (5.0s) 15:13:45 ~ 15:13:55 (5.1s) 15:14:58 ~ 15:15:08 (5.1s) 15:14:34 ~ 15:14:39 (2.8s) 15:14:53 ~ 15:14:58 (4.2s)

提示词

查看本步 Prompt

Knowledge cutoff date? Reply with YYYY-MM only.

HTTP 头(TTFB) 4,174 ms 3,948 ms 3,339 ms 1,632 ms 2,686 ms

首文本(SSE delta) 4,175 ms 3,949 ms 3,340 ms 1,781 ms -

总耗时 4,270 ms 4,015 ms 4,253 ms 1,813 ms 3,108 ms

输入 3 3 8 3 3

输出 8 8 39 8 8

缓存创建

5m: 0 / 1h: 38

33,966

5m: 33,966 / 1h: 0

5m: 0 / 1h: 38

5m: 0 / 1h: 57

缓存读取 32,520 32,521 0 32,520 32,659

输出速度 - - 42.7 tok/s - -

响应

2025-08

2025-08

2025-08

2025-08

2025-08

步骤 4: identity_free text

时间 15:15:58 ~ 15:16:08 (5.0s) 15:15:49 ~ 15:15:59 (7.7s) 15:18:40 ~ 15:18:50 (5.4s) 15:16:59 ~ 15:17:04 (4.4s) 15:16:31 ~ 15:16:41 (9.4s)

提示词

查看本步 Prompt

用中文回答：你是谁？由谁开发？50 字以内。

HTTP 头(TTFB) 4,225 ms 3,944 ms 2,693 ms 2,664 ms 4,253 ms

首文本(SSE delta) 4,226 ms 3,945 ms 2,694 ms 3,054 ms -

总耗时 4,316 ms 5,926 ms 4,625 ms 3,410 ms 7,900 ms

输入 3 3 9 3 3

输出 55 46 57 48 51

缓存创建

5m: 0 / 1h: 35

5m: 0 / 1h: 39

34,342

5m: 34,342 / 1h: 0

5m: 0 / 1h: 41

缓存读取 32,558 32,559 0 32,558 32,716

输出速度 - 23.2 tok/s 29.5 tok/s 134.8 tok/s 14.0 tok/s

响应

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。

我是 Claude，由 Anthropic 公司打造的 AI 助手，目前运行的模型版本为 claude-sonnet-4-6。

我是 Claude，由 Anthropic 打造的 AI 助手，当前运行的模型版本为 claude-sonnet-4-6。

我叫 Claude，是由 Anthropic 公司开发的 AI 助手，当前使用的模型版本为 claude-sonnet-4-6。

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。

步骤 5: knowledge_recall text

时间 15:17:10 ~ 15:17:15 (4.5s) 15:18:05 ~ 15:18:15 (7.3s) 15:22:13 ~ 15:22:43 (26.6s) 15:20:29 ~ 15:20:34 (4.4s) 15:19:46 ~ 15:19:51 (4.8s)

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) 3,633 ms 2,724 ms 1,657 ms 1,524 ms 1,822 ms

首文本(SSE delta) 3,634 ms 2,726 ms 1,658 ms 1,526 ms -

总耗时 3,693 ms 4,958 ms 25,720 ms 3,374 ms 3,735 ms

输入 3 3 55 3 3

输出 204 204 2,100 205 205

缓存创建

8,037

5m: 0 / 1h: 8,037

8,033

5m: 0 / 1h: 8,033

9,663

5m: 9,663 / 1h: 0

8,036

5m: 0 / 1h: 8,036

8,058

5m: 0 / 1h: 8,058

缓存读取 0 0 0 0 0

输出速度 burst 91.4 tok/s 87.3 tok/s 110.9 tok/s 107.2 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

步骤 6: digit_count text

时间 15:21:07 ~ 15:21:22 (12.3s) 15:19:54 ~ 15:20:19 (22.5s) 15:25:15 ~ 15:25:40 (21.2s) 15:22:49 ~ 15:23:09 (15.3s) 15:22:51 ~ 15:23:16 (21.3s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=419453226854967872263455395035517875244202741931325750178645549742539755

HTTP 头(TTFB) 6,538 ms 3,018 ms 2,206 ms 1,459 ms 6,254 ms

首文本(SSE delta) 6,539 ms 3,402 ms 2,207 ms 1,841 ms -

总耗时 11,583 ms 20,305 ms 20,335 ms 14,329 ms 20,118 ms

输入 164 164 93 164 3

输出 823 1,295 2,261 997 1,600

缓存创建

5m: 0 / 1h: 0

116

5m: 116 / 1h: 0

5m: 0 / 1h: 0

35,752

5m: 0 / 1h: 35,752

缓存读取 0 0 0 0 0

输出速度 burst 76.6 tok/s 124.7 tok/s 79.8 tok/s 115.4 tok/s

响应

```json
{"n":72,"c":[3,9,6]}
```

## Counting digits in s

**s** = `419453226854967872263455395035517875244202741931325750178645549742539755`

### Length
Grouping into blocks of 10:
`4194532268` `5496787226` `3455395035` `5178752442`

Counting `s = 419453226854967872263455395035517875244202741931325750178645549742539755`:

**Total length (n):** 72

**0s:** positions 28, 41, 54 → **3**
**7s:** positions 15, 17, 33, 35, 43, 52, 56, 6

## Counting digits in s

**s** = `419453226854967872263455395035517875244202741931325750178645549742539755`

### Length
Splitting into groups of 10:
`4194532268` `5496787226` `3455395035` `5178752442`

Counting character by character:

```
Pos:  1  2  3  4  5  6  7  8  9 10 11 12 13 14 15 16 17 18 19 20
Char: 4  1  9  4  5  3  2  2  6  8  5  4  9  6  7  8  7  2  2  6

Pos: 21 22 23 24 25 26 27 28 29

总计

总输入(含缓存) 138,205 138,209 144,815 138,210 174,417

总输出 1,117 1,580 4,556 1,285 1,891

总缓存创建

40,630

5m: 0 / 1h: 40,630

40,631

5m: 0 / 1h: 40,631

111,578

5m: 111,578 / 1h: 0

40,635

5m: 0 / 1h: 40,635

76,567

5m: 0 / 1h: 76,567

总缓存读取 95,932 97,399 32,958 97,396 97,832

输出速度 burst 73.2 tok/s 96.4 tok/s 85.1 tok/s 95.6 tok/s

总执行时间 35.2s 56.3s 1m 7.1s 35.1s 50.8s

总墙钟时间

11m 22.0s

含步骤间等待

10m 17.0s

含步骤间等待

15m 36.1s

含步骤间等待

13m 2.9s

含步骤间等待

13m 18.0s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.2028

Anthropic: Claude Sonnet 4.6

$0.2058

Anthropic: Claude Sonnet 4.6

$0.3406

Anthropic: Claude Sonnet 4.6

$0.2014

Anthropic: Claude Sonnet 4.6

$0.3449

Anthropic: Claude Sonnet 4.6

通道指纹

来源步骤 digit_count 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 (100%) Anthropic 直连 (100%) Anthropic 直连 (100%) Anthropic 直连 (100%)

CDN - cloudflare cloudflare - -

ID 格式 msg_01JfcbYtYFitiYgHdBYe1kUD msg_01RVPJBNHFpTN77RDntcU5DX msg_01DVcKm0ZtKd3Bh29zkZrlBt msg_01BWhAB17zvuCLNZ2owDLCVJ msg_01StTWg5VbUBtj9iNYPoUm5q

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	server: nginx/1.27.5	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	tr-ew7y4grd-k7ftlc0abs-187789
O-Max	cf-ray: a1332a752ef08eab-AMS server: cloudflare x-oneapi-request-id: 20260629151957544711237iAPmptoU	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	a1332a752ef08eab-AMS
R-super	cf-ray: a13332427f07734b-NRT server: cloudflare x-new-api-version: v0.0.0 x-oneapi-request-id: 202606290725167319906148268d9d6KVUu1UBq	id_format: msg_prefix 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-idhdr:server:cloudflarehdr:cf-raybody:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsupstream:anthropic-direct-from-body	a13332427f07734b-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_011CcX9chjgrTbNYAJDDmuou
O-Max	server: nginx/1.29.8 x-oneapi-request-id: 202606290722529078710018268d9d6pQx6IILQ	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-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 94 96 68 100 97

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 0.0 /10 ×20 10.0 /10 ×20 10.0 /10 ×20

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

缓存连续性 10.0 /10 ×14 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 0.0 /10 ×13 10.0 /10 ×13 10.0 /10 ×13

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

系统提示纯净 10.0 /10 ×8 10.0 /10 ×8 5.0 /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

知识截止 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

service_tier 10.0 /10 ×6 10.0 /10 ×6 10.0 /10 ×6 10.0 /10 ×6 10.0 /10 ×6

inference_geo 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5

延迟基线 9.8 /10 ×5 8.4 /10 ×5 9.5 /10 ×5 10.0 /10 ×5 8.5 /10 ×5

流式投递 0.0 /10 ×5 10.0 /10 ×5 9.6 /10 ×5 10.0 /10 ×5 未计入

Req-ID 透传 0.0 /10 ×4 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 10.0 /10 ×3

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

综合结论

与基线相似度 94 v3.25.0 96 v3.25.0 68 v3.25.0 基准（参考） 97 v3.25.0

雷达图例

共 20 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 流式投递 buffer_dump_match
5 自由身份 identity_free_clean
6 停止原因 stop_reason_present
7 系统提示纯净 system_prompt_clean
8 Service Tier service_tier_present
9 缓存命中比 cache_hit_ratio_match
10 缓存 TTL 一致 cache_ttl_consistency
11 推理地区 inference_geo_present
12 缓存连续 cache_continuity_intra
13 知识召回 knowledge_recall_match
14 延迟基线 latency_baseline_match
15 原生 msg-ID anthropic_msg_id_format
16 Sliding 缓存 cache_sliding_correctness
17 结构化身份 identity_structured_match
18 信封自报一致 envelope_self_report_match
19 档位思考量 tier_thinking_volume_match
20 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 5114 ms 7609 ms 10335 ms 4867 ms 7235 ms

输出速度 burst 73.2 tok/s 96.4 tok/s 85.1 tok/s 95.6 tok/s

完成步骤 6/6 成功 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 ⓘ

$0.2028

$0.2058

$0.3406

$0.2014

$0.3449

缓存读取占比 ⓘ 69% 70% 23% 70% 56%

揭晓

通道

服务商 TopRouterCN ↗

通道O-Max

服务商 ClaudeCN ↗

通道O-Max

服务商 FoxCode ↗

通道R-super

官方基准

服务商 ByteCatCode ↗

通道O-Max

返回列表