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

2026/06/19 运行日期

claude-sonnet-4-6 官方基准

2 通道分数 < 70 与基线相似度

已揭晓揭晓进度

步骤 / 指标

LinkAPI

O-Max

官方基准

WorldBase.ai

M-

SSSAiCode

R-Kiro

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 10:05:24 ~ 10:05:29 (4.4s) 10:05:32 ~ 10:05:37 (4.1s) 10:05:28 ~ 10:05:33 (4.9s) 10:05:30 ~ 10:05:35 (3.8s) 10:05:26 ~ 10:05:36 (6.8s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 3,140 ms 2,636 ms 3,128 ms 2,633 ms 5,146 ms

首文本(SSE delta) 3,144 ms 2,638 ms 3,169 ms 2,673 ms -

总耗时 3,272 ms 2,774 ms 3,268 ms 2,777 ms 5,291 ms

输入 3 3 0 0 3

输出 5 5 1 2 5

缓存创建

32,318

5m: 0 / 1h: 32,318

32,318

5m: 0 / 1h: 32,318

35,312

5m: 0 / 1h: 35,312

32,256

5m: 0 / 1h: 32,256

缓存读取 0 0 0 0 0

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 10:08:36 ~ 10:08:41 (4.5s) 10:07:48 ~ 10:07:53 (3.8s) 10:06:53 ~ 10:07:03 (6.5s) 10:07:53 ~ 10:07:58 (3.3s) 10:08:00 ~ 10:08:10 (5.6s)

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) 2,843 ms 2,449 ms 3,954 ms 2,348 ms 3,750 ms

首文本(SSE delta) 2,844 ms 2,451 ms 3,955 ms 2,353 ms -

总耗时 3,323 ms 2,837 ms 4,423 ms 2,608 ms 4,194 ms

输入 3 3 214 0 3

输出 22 22 21 17 22

缓存创建

191

5m: 0 / 1h: 191

191

5m: 0 / 1h: 191

5m: 0 / 1h: 1

191

5m: 0 / 1h: 191

缓存读取 32,318 32,318 35,293 0 32,256

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 10:11:09 ~ 10:11:14 (4.5s) 10:09:45 ~ 10:09:50 (4.1s) 10:08:59 ~ 10:09:09 (5.3s) 10:11:55 ~ 10:12:00 (3.5s) 10:11:43 ~ 10:11:53 (6.7s)

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) 2,856 ms 2,991 ms 3,052 ms 2,333 ms 4,684 ms

首文本(SSE delta) 3,218 ms 2,993 ms 3,053 ms 2,593 ms -

总耗时 3,407 ms 3,134 ms 4,071 ms 2,596 ms 5,642 ms

输入 3 3 37 0 3

输出 8 8 4 3 8

缓存创建

5m: 0 / 1h: 55

5m: 0 / 1h: 38

239

5m: 0 / 1h: 239

5m: 0 / 1h: 38

缓存读取 32,509 32,509 35,285 0 32,447

响应

2025-08

2025-08

2025-08

2025-08

2025-08

步骤 4: identity_free text

时间 10:12:32 ~ 10:12:42 (4.9s) 10:12:30 ~ 10:12:35 (4.8s) 10:10:41 ~ 10:10:51 (5.3s) 10:14:36 ~ 10:14:46 (5.5s) 10:13:54 ~ 10:14:04 (5.4s)

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) 3,173 ms 2,967 ms 2,900 ms 4,254 ms 3,291 ms

首文本(SSE delta) 3,174 ms 2,968 ms 2,901 ms 4,255 ms -

总耗时 3,868 ms 3,686 ms 4,186 ms 4,764 ms 4,264 ms

输入 3 3 33 0 3

输出 62 44 58 24 55

缓存创建

5m: 0 / 1h: 43

5m: 0 / 1h: 39

5m: 0 / 1h: 41

5m: 0 / 1h: 35

缓存读取 32,564 32,547 35,522 0 32,485

输出速度 89.3 tok/s 61.3 tok/s 45.1 tok/s - 56.5 tok/s

响应

我是 Claude，由 Anthropic 开发的 AI 助手。我能帮助完成写作、分析、编程、问答等多种任务，致力于提供安全、有益的服务。

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。

我是 Kiro，一款由 Anthropic 旗下技术驱动、基于 Claude 模型构建的 AI 开发环境助手，帮助开发者更高效地完成软件工程任务。

我叫 Claude，是由 Anthropic 开发的 AI 助手。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 10:16:41 ~ 10:16:51 (7.9s) 10:15:33 ~ 10:15:38 (4.4s) 10:13:18 ~ 10:13:28 (6.3s) 10:18:47 ~ 10:18:57 (5.0s) 10:17:43 ~ 10:17:53 (6.5s)

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) 4,691 ms 1,561 ms 2,172 ms 1,871 ms 3,660 ms

首文本(SSE delta) 4,702 ms 1,562 ms 2,173 ms 1,948 ms -

总耗时 6,876 ms 3,399 ms 5,091 ms 3,867 ms 5,376 ms

输入 2 3 1,022 0 3

输出 317 204 183 221 198

缓存创建

10,916

5m: 0 / 1h: 10,916

8,019

5m: 0 / 1h: 8,019

8,539

5m: 0 / 1h: 8,539

32,076

5m: 0 / 1h: 32,076

缓存读取 0 0 0 0 0

输出速度 145.8 tok/s 111.1 tok/s 62.7 tok/s 115.2 tok/s 115.4 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

步骤 6: digit_count text

时间 10:19:15 ~ 10:19:35 (15.8s) 10:19:09 ~ 10:19:29 (16.0s) 10:15:58 ~ 10:16:23 (20.8s) 10:20:49 ~ 10:21:09 (17.1s) 10:19:33 ~ 10:19:53 (17.3s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=553373841223457803283855714445905714021590864074368685217698265736729088

HTTP 头(TTFB) 1,163 ms 6,342 ms 3,689 ms 1,678 ms 3,744 ms

首文本(SSE delta) 1,164 ms 6,344 ms 3,690 ms 1,719 ms -

总耗时 14,661 ms 15,049 ms 19,239 ms 16,359 ms 16,206 ms

输入 164 164 203 0 3

输出 1,459 926 14 1,097 1,295

缓存创建

5m: 0 / 1h: 0

36,099

5m: 0 / 1h: 36,099

缓存读取 0 0 0 0 0

输出速度 108.1 tok/s 106.4 tok/s - 74.9 tok/s 103.9 tok/s

响应

## Counting digits in s

**s = `553373841223457803283855714445905714021590864074368685217698265736729088`**

### Step 1 – Length

Grouping in tens:
```
5533738412  (10)
2345780328  (20)
3855714445  (3

```json
{"n":72,"c":[6,8,4]}
```

**Verification:**

String: `553373841223457803283855714445905714021590864074368685217698265736729088`

- **Length**: 72
- **0s** (×6): positions 17, 32, 37, 42, 46, 7


{"n":72,"c":[6,8,4]}

```
s = "553373841223457803283855714445905714021590864074368685217698265736729088"
```

Counting position by position:

Total length n = 72

0s at positions: 17, 32, 37, 42, 46, 70 → **6**
7s at posit

Counting character by character:

```
s = 553373841223457803283855714445905714021590864074368685217698265736729088
```

**Length — n:**
```
5533738412234578 0328385571444590 5714021590864074 368685217

总计

总输入(含缓存) 141,092 138,158 151,741 0 197,901

总输出 1,873 1,209 281 1,364 1,583

总缓存创建

43,523

5m: 0 / 1h: 43,523

40,605

5m: 0 / 1h: 40,605

44,132

5m: 0 / 1h: 44,132

100,695

5m: 0 / 1h: 100,695

总缓存读取 97,391 97,374 106,100 0 97,188

输出速度 112.3 tok/s 104.3 tok/s 57.3 tok/s 79.6 tok/s 102.2 tok/s

总执行时间 41.9s 37.2s 49.2s 38.3s 48.2s

总墙钟时间

14m 11.9s

含步骤间等待

13m 57.9s

含步骤间等待

10m 55.1s

含步骤间等待

15m 39.3s

含步骤间等待

14m 27.1s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.2211

Anthropic: Claude Sonnet 4.6

$0.2002

Anthropic: Claude Sonnet 4.6

$0.2061

Anthropic: Claude Sonnet 4.6

>=$0.0205

Anthropic: Claude Sonnet 4.6 (成本不完整)

$0.4306

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 one-api (100%) direct (100%) direct (100%) direct (75%) one-api (100%)

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

CDN - - cloudflare - -

ID 格式 msg_014PfQ8xCk13jonoebMshs4K msg_01RyEXB47udAs9nAkKQPS72Y msg_01qrB5Z88W1w7o1RfG6pk18v msg_2a4f8194ac6046d596fbe89108773c19 msg_019z9yiTXwLDWGgwDDLjYMNK

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 set-cookie: server_name_session=9a42ff809bdc7f49eaa4735d20f97eff; Max-Age=86400; httponly; path=/ x-new-api-version: v1.0.0-rc.11 x-oneapi-request-id: 20260619021917796355708268d9d6Mh2wszzD	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	—
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_011CcBpNVZHBDYXygT8LZh8F
M-	cf-ray: a0df07761c176716-AMS server: cloudflare	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:server:cloudflarehdr:cf-raybody:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	c85f776d-5d85-408b-9bdf-60cc315ac94f a0df07761c176716-AMS
R-Kiro	server: Photon-Edge x-amzn-requestid: 64e66a3c-2916-46d8-9dd9-237e45f5161d	id_format: msg_prefix has_service_tier: false has_inference_geo: false model_has_date_suffix: false usage_has_cache_fields: true	body:id:msg_prefixbody:usage.cache_fieldsupstream:x-amzn-requestidupstream:anthropic-direct-from-body	1781835650230-8xjrxm0v9by9o 64e66a3c-2916-46d8-9dd9-237e45f5161d
O-Max	server: nginx/1.29.8 x-oneapi-request-id: 202606190219342410240198268d9d6gjKZ9fVB	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 93 96 81 57 89

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 6.8 /10 ×20 7.3 /10 ×20 7.1 /10 ×20 0.0 /10 ×20 3.7 /10 ×20

缓存 TTL 一致 10.0 /10 ×15 10.0 /10 ×15 10.0 /10 ×15 0.0 /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 5.0 /10 ×14 10.0 /10 ×14

sliding 5m 10.0 /10 ×13 10.0 /10 ×13 10.0 /10 ×13 5.0 /10 ×13 10.0 /10 ×13

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

系统提示纯净 10.0 /10 ×8 10.0 /10 ×8 0.0 /10 ×8 未计入 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 0.0 /10 ×7 10.0 /10 ×7 10.0 /10 ×7

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

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

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

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

Req-ID 透传 0.0 /10 ×4 10.0 /10 ×4 0.0 /10 ×4 0.0 /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 93 v3.24.0 基准（参考） 81 v3.24.0 57 v3.24.0 89 v3.24.0

雷达图例

共 21 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 中文遵循 instruction_following_lang
20 档位思考量 tier_thinking_volume_match
21 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 5901 ms 5147 ms 6713 ms 5495 ms 6829 ms

输出速度 112.3 tok/s 104.3 tok/s 57.3 tok/s 79.6 tok/s 102.2 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 ⓘ

$0.2211

$0.2002

$0.2061

>=$0.0205 (成本不完整)

$0.4306

缓存读取占比 ⓘ 69% 70% 70% - 49%

揭晓

通道

服务商 LinkAPI ↗

通道O-Max

官方基准

服务商 WorldBase.ai ↗

通道M-

服务商 SSSAiCode ↗

通道R-Kiro

服务商 ByteCatCode ↗

通道O-Max

返回列表