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

2026/06/20 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

luminai

M-Kiro/Antigravity

FastCode

PackyCode

cc-Zeus

duckcoding

cc专用

官方基准

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 10:44:19 ~ 10:44:29 (6.2s) 10:44:21 ~ 10:44:31 (5.6s) 10:44:17 ~ 10:44:22 (3.8s) 10:44:15 ~ 10:44:25 (6.0s) 10:44:23 ~ 10:44:28 (4.3s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 4,563 ms 3,625 ms 2,291 ms 4,253 ms 3,016 ms

首文本(SSE delta) 4,630 ms 3,627 ms 2,293 ms 4,255 ms 3,030 ms

总耗时 4,688 ms 3,984 ms 2,457 ms 4,696 ms 3,191 ms

输入 1,443 3 3 3 3

输出 1 5 5 5 5

缓存创建

4,111

5m: 4,111 / 1h: 0

32,318

5m: 0 / 1h: 32,318

32,318

5m: 0 / 1h: 32,318

32,318

5m: 0 / 1h: 32,318

32,318

5m: 0 / 1h: 32,318

缓存读取 23,298 0 0 0 0

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 10:47:25 ~ 10:47:35 (5.6s) 10:47:31 ~ 10:47:41 (5.7s) 10:45:38 ~ 10:45:43 (3.7s) 10:47:45 ~ 10:47:50 (4.7s) 10:47:23 ~ 10:47:28 (3.7s)

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) 4,040 ms 3,744 ms 2,707 ms 2,936 ms 2,228 ms

首文本(SSE delta) 4,041 ms 3,999 ms 2,708 ms 2,939 ms 2,229 ms

总耗时 4,277 ms 4,000 ms 2,709 ms 3,762 ms 2,663 ms

输入 1,452 3 3 3 3

输出 24 22 22 22 22

缓存创建

4,137

5m: 4,137 / 1h: 0

32,509

5m: 0 / 1h: 32,509

191

5m: 0 / 1h: 191

191

5m: 0 / 1h: 191

191

5m: 0 / 1h: 191

缓存读取 23,444 0 32,318 32,318 32,318

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 10:49:07 ~ 10:49:17 (5.7s) 10:49:55 ~ 10:50:05 (5.6s) 10:48:48 ~ 10:48:53 (3.1s) 10:50:37 ~ 10:50:52 (10.8s) 10:49:10 ~ 10:49:15 (3.3s)

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) 4,220 ms 4,369 ms 1,856 ms 9,231 ms 1,928 ms

首文本(SSE delta) 4,484 ms 4,370 ms 1,857 ms 9,350 ms 2,075 ms

总耗时 4,548 ms 4,370 ms 2,065 ms 9,696 ms 2,137 ms

输入 1,455 3 3 3 3

输出 4 8 8 8 8

缓存创建

4,146

5m: 4,146 / 1h: 0

32,547

5m: 0 / 1h: 32,547

5m: 0 / 1h: 38

5m: 0 / 1h: 57

5m: 0 / 1h: 55

缓存读取 23,496 0 32,509 32,509 32,509

响应

2025-08

2025-08

2025-08

2025-08

2025-08

步骤 4: identity_free text

时间 10:50:33 ~ 10:50:43 (7.3s) 10:52:14 ~ 10:52:24 (5.4s) 10:52:04 ~ 10:52:14 (5.1s) 10:52:17 ~ 10:52:32 (12.6s) 10:53:07 ~ 10:53:12 (4.3s)

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) 4,034 ms 4,088 ms 2,237 ms 10,410 ms 2,050 ms

首文本(SSE delta) 4,787 ms 4,089 ms 2,238 ms 10,411 ms 2,052 ms

总耗时 5,735 ms 4,146 ms 3,428 ms 11,528 ms 3,046 ms

输入 1,458 3 3 3 3

输出 48 46 69 49 63

缓存创建

4,152

5m: 4,152 / 1h: 0

268

5m: 0 / 1h: 268

5m: 0 / 1h: 43

缓存读取 23,532 32,318 32,547 32,566 32,564

输出速度 50.6 tok/s - 58.0 tok/s 43.9 tok/s 63.4 tok/s

响应

我是 Claude，由 Anthropic 开发的 AI 助手。当前版本为 Claude Sonnet 4.6，擅长代码、写作、分析等多类任务。

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 开发的 AI 助手。我能帮助您完成写作、分析、编程、问答等各类任务。当前运行的模型版本是 claude-sonnet-4-6。

我是 Claude，由 Anthropic 开发的 AI 助手。我能帮助你完成编程、写作、分析、问答等各类任务。

我是 Claude，由 Anthropic 开发的 AI 助手。我能帮助你完成写作、分析、编程、问答等各类任务，致力于提供安全、有益的回应。

步骤 5: knowledge_recall text

时间 10:53:02 ~ 10:53:12 (6.1s) 10:54:11 ~ 10:54:21 (6.1s) 10:55:28 ~ 10:55:38 (6.5s) 10:53:43 ~ 10:53:53 (5.2s) 10:54:56 ~ 10:55:06 (4.4s)

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) 2,747 ms 2,628 ms 2,537 ms 2,107 ms 1,480 ms

首文本(SSE delta) 4,632 ms 2,629 ms 2,539 ms 2,108 ms 1,481 ms

总耗时 4,713 ms 4,878 ms 5,091 ms 4,320 ms 3,483 ms

输入 375 3 3 3 3

输出 210 204 205 205 205

缓存创建

1,067

5m: 1,067 / 1h: 0

8,021

5m: 0 / 1h: 8,021

8,049

5m: 0 / 1h: 8,049

8,047

5m: 0 / 1h: 8,047

8,059

5m: 0 / 1h: 8,059

缓存读取 6,050 0 0 0 0

输出速度 burst 90.7 tok/s 80.3 tok/s 92.7 tok/s 102.4 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 10:55:43 ~ 10:55:53 (5.4s) 10:56:28 ~ 10:56:49 (15.3s) 10:59:14 ~ 10:59:34 (16.8s) 10:55:25 ~ 10:55:41 (14.7s) 10:56:10 ~ 10:56:30 (17.3s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=825689387103819722820396671244224913499205385655551538234747020911963381

HTTP 头(TTFB) 3,891 ms 1,951 ms 1,197 ms 1,112 ms 1,107 ms

首文本(SSE delta) 3,909 ms 1,952 ms 1,591 ms 1,114 ms 1,512 ms

总耗时 3,990 ms 14,108 ms 15,791 ms 13,740 ms 16,408 ms

输入 190 164 164 164 164

输出 14 1,172 1,185 1,197 1,259

缓存创建

5m: 0 / 1h: 0

缓存读取 - 0 0 0 0

输出速度 - 96.4 tok/s 83.5 tok/s 94.8 tok/s 84.5 tok/s

响应

{"n":71,"c":[3,4,5]}

## Counting digits in s

**s = 825689387103819722820396671244224913499205385655551538234747020911963381**

### Splitting into groups of 10:
```
8256893871  (1–10)
0381972282  (11–20)
0396671244  (21–3

## Counting digits in `s`

**s = `825689387103819722820396671244224913499205385655551538234747020911963381`**

### Step 1 – Length

Breaking into groups of 10:
```
8256893871  → 10
0381972282  → 10
03

## Counting digits in s

**s = 825689387103819722820396671244224913499205385655551538234747020911963381**

### Length
Grouping in 10s:
```
8256893871 | 0381972282 | 0396671244 | 2249134992 | 053856555

## Counting digits in s

**s** = `825689387103819722820396671244224913499205385655551538234747020911963381`

### Length

Splitting into groups of 10:
```
8256893871 | 0381972282 | 0396671244 | 2249134

总计

总输入(含缓存) 123,806 138,160 138,192 138,228 138,236

总输出 301 1,457 1,494 1,486 1,562

总缓存创建

17,613

5m: 17,613 / 1h: 0

105,663

5m: 0 / 1h: 105,663

40,639

5m: 0 / 1h: 40,639

40,656

5m: 0 / 1h: 40,656

40,666

5m: 0 / 1h: 40,666

总缓存读取 99,820 32,318 97,374 97,393 97,391

输出速度 50.6 tok/s 95.5 tok/s 81.3 tok/s 90.9 tok/s 85.3 tok/s

总执行时间 36.3s 43.7s 38.8s 53.9s 37.4s

总墙钟时间

11m 34.1s

含步骤间等待

12m 27.9s

含步骤间等待

15m 17.3s

含步骤间等待

11m 26.0s

含步骤间等待

12m 7.7s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.0949

Anthropic: Claude Sonnet 4.6

$0.4283

Anthropic: Claude Sonnet 4.6

$0.2046

Anthropic: Claude Sonnet 4.6

$0.2045

Anthropic: Claude Sonnet 4.6

$0.2057

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

ID 格式 msg_01DtiCM282bRR2pBkuJXSLAjPX msg_vrtx_015opkWhWLCRVN4kKdd7PrUs msg_0176gfLjshLtVVVEDKJU2YJq msg_0126qgtGfQFnwu9jjpzj2p9B msg_01YSeyats8Btp9aVoK8PXarG

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
M-Kiro/Antigravity	cf-ray: a0e77f0f7b7a0eab-AMS server: cloudflare x-oneapi-request-id: 20260620025545382452358268d9d64vF194ET	id_format: msg_prefix extra_fields: ["container"] has_service_tier: true has_inference_geo: false 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.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	a0e77f0f7b7a0eab-AMS
cc	server: nginx	id_format: msg_prefix extra_fields: ["stop_details"] has_service_tier: false has_inference_geo: false model_has_date_suffix: false usage_has_cache_fields: true	body:id:msg_prefixbody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	—
cc-Zeus	via: 1.1 Caddy cf-ray: a0e784348f827650-LAX server: cloudflare x-oneapi-request-id: 2026062010591530874497492vi18Zv	id_format: msg_prefix via_chain: ["1.1 Caddy"] 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-rayhdr:viabody:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	a0e784348f827650-LAX
cc专用	server: Nginx x-oneapi-request-id: 202606200255271263281598268d9d6al3aJYAv	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	—
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_011CcDm172H6sPLfPLnpkRnw

协议指纹评分（机器自动）

总分 / 100 74 71 98 96 100

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

缓存 TTL 一致 0.0 /10 ×15 10.0 /10 ×15 10.0 /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 5.3 /10 ×13 10.0 /10 ×13 10.0 /10 ×13 10.0 /10 ×13

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

系统提示纯净 0.0 /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

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

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

延迟基线 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 5.0 /10 ×5 10.0 /10 ×5

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

Req-ID 透传 0.0 /10 ×4 0.0 /10 ×4 0.0 /10 ×4 0.0 /10 ×4 10.0 /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 74 v3.24.0 71 v3.24.0 98 v3.24.0 96 v3.24.0 基准（参考）

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 4659 ms 5914 ms 5257 ms 7957 ms 5155 ms

输出速度 50.6 tok/s 95.5 tok/s 81.3 tok/s 90.9 tok/s 85.3 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 ⓘ

$0.0949

$0.4283

$0.2046

$0.2045

$0.2057

缓存读取占比 ⓘ 81% 23% 70% 70% 70%

揭晓

通道

服务商 luminai ↗

通道M-Kiro/Antigravity

服务商 FastCode ↗

通道cc

服务商 PackyCode ↗

通道cc-Zeus

服务商 duckcoding ↗

通道cc专用

官方基准

返回列表