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

2026/07/02 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

官方基准

SSSAiCode

R-Kiro

sucui

O-Max

PackyCode

cc-Zeus

AIMZ

M-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 v2.1.195 claude-cli v2.1.195 claude-cli v2.1.195 claude-cli v2.1.195 claude-cli v2.1.195

偏差标红基于官方基准

步骤 1: ping text

时间 16:01:59 ~ 16:02:04 (3.4s) 16:01:53 ~ 16:01:58 (3.5s) 16:01:55 ~ 16:02:05 (5.2s) 16:01:51 ~ 16:01:56 (4.2s) 16:01:57 ~ 16:02:02 (3.8s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 2,203 ms 2,677 ms 3,977 ms 1,934 ms 2,226 ms

首文本(SSE delta) 2,204 ms 2,797 ms 3,978 ms 1,935 ms 2,320 ms

总耗时 2,331 ms 2,797 ms 3,979 ms 2,240 ms 2,740 ms

输入 3 0 3 3 3

输出 5 2 5 5 5

缓存创建

5,991

5m: 0 / 1h: 5,991

5,991

5m: 0 / 1h: 5,991

5,991

5m: 0 / 1h: 5,991

20,511

5m: 0 / 1h: 20,511

缓存读取 14,884 0 14,884 14,884 0

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 16:05:16 ~ 16:05:21 (3.1s) 16:03:03 ~ 16:03:08 (4.4s) 16:04:29 ~ 16:04:34 (4.4s) 16:05:33 ~ 16:05:38 (4.2s) 16:05:37 ~ 16:05:48 (6.5s)

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) 1,634 ms 3,312 ms 2,214 ms 1,831 ms 5,035 ms

首文本(SSE delta) 1,636 ms 3,415 ms 2,298 ms 1,832 ms 5,463 ms

总耗时 2,113 ms 3,715 ms 3,368 ms 3,033 ms 5,463 ms

输入 3 0 3 3 3

输出 22 18 22 22 22

缓存创建

202

5m: 0 / 1h: 202

202

5m: 0 / 1h: 202

202

5m: 0 / 1h: 202

202

5m: 0 / 1h: 202

缓存读取 20,875 0 20,875 20,875 20,511

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-4-5

步骤 3: cutoff text

时间 16:08:14 ~ 16:08:19 (4.8s) 16:04:49 ~ 16:04:54 (2.9s) 16:07:10 ~ 16:07:21 (6.0s) 16:08:50 ~ 16:08:55 (3.5s) 16:08:54 ~ 16:09:04 (5.0s)

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) 3,546 ms 1,809 ms 2,525 ms 2,166 ms 2,619 ms

首文本(SSE delta) 3,730 ms 2,100 ms 3,707 ms 2,170 ms 2,769 ms

总耗时 3,873 ms 2,240 ms 3,941 ms 2,432 ms 2,799 ms

输入 3 0 3 3 3

输出 8 3 8 8 8

缓存创建

5m: 0 / 1h: 57

5m: 0 / 1h: 55

缓存读取 21,077 0 21,077 21,077 20,713

响应

2025-08

2025-08

2025-08

2025-08

2025-04

步骤 4: identity_free text

时间 16:10:24 ~ 16:10:39 (4.0s) 16:07:23 ~ 16:07:28 (3.9s) 16:08:29 ~ 16:08:39 (5.4s) 16:12:39 ~ 16:12:49 (5.6s) 16:13:03 ~ 16:13:08 (4.0s)

提示词

查看本步 Prompt

用中文说说你的身份：什么产品？由谁打造？50 字以内。

HTTP 头(TTFB) 1,830 ms 1,756 ms 3,182 ms 2,599 ms 2,498 ms

首文本(SSE delta) 1,832 ms 2,016 ms 3,845 ms 2,606 ms 2,696 ms

总耗时 2,717 ms 3,134 ms 4,345 ms 3,967 ms 2,802 ms

输入 3 0 3 3 3

输出 46 36 66 44 24

缓存创建

5m: 0 / 1h: 39

5m: 0 / 1h: 43

5m: 0 / 1h: 39

5m: 0 / 1h: 35

缓存读取 21,134 0 21,132 21,132 20,768

输出速度 52.0 tok/s 32.2 tok/s 132.0 tok/s 32.3 tok/s -

响应

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

我是 Kiro，一款由 Anthropic 开发的 AI 驱动开发环境，基于 Claude 模型构建，帮助开发者专注于系统设计和决策，而非繁琐的编码工作。

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、有益的智能对话服务。

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 Claude，由 Anthropic 开发的 AI 助手。

步骤 5: knowledge_recall text

时间 16:12:50 ~ 16:13:00 (5.5s) 16:08:56 ~ 16:09:01 (4.9s) 16:10:34 ~ 16:10:44 (8.6s) 16:15:24 ~ 16:15:34 (5.8s) 16:14:51 ~ 16:15:01 (6.0s)

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) 2,773 ms 2,083 ms 4,360 ms 2,319 ms 2,483 ms

首文本(SSE delta) 2,775 ms 2,258 ms 4,361 ms 2,320 ms 4,814 ms

总耗时 4,562 ms 4,119 ms 7,346 ms 4,787 ms 4,961 ms

输入 3 0 3 3 3

输出 205 223 204 205 198

缓存创建

6,676

5m: 0 / 1h: 6,676

6,698

5m: 0 / 1h: 6,698

6,672

5m: 0 / 1h: 6,672

8,415

5m: 0 / 1h: 8,415

缓存读取 2,131 0 2,131 2,131 0

输出速度 114.7 tok/s 119.8 tok/s 68.3 tok/s 83.1 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 16:15:20 ~ 16:15:40 (18.4s) 16:10:52 ~ 16:11:07 (14.3s) 16:14:21 ~ 16:14:36 (13.3s) 16:16:40 ~ 16:16:55 (13.0s) 16:18:30 ~ 16:19:00 (28.0s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=476423909362133322888007718815186500190932742072030926205404904190112781

HTTP 头(TTFB) 1,455 ms 1,881 ms 1,035 ms 1,151 ms 1,725 ms

首文本(SSE delta) 1,457 ms 1,917 ms 1,036 ms 1,152 ms 2,148 ms

总耗时 17,355 ms 13,582 ms 12,160 ms 11,886 ms 26,995 ms

输入 164 0 164 164 110

输出 1,565 812 979 926 2,092

缓存创建

5m: 0 / 1h: 0

缓存读取 0 0 0 0 0

输出速度 98.4 tok/s 69.6 tok/s 88.0 tok/s 86.3 tok/s 84.2 tok/s

响应

```
s = "476423909362133322888007718815186500190932742072030926205404904190112781"
```

**Parsing digit by digit (indexed 1–72):**

| pos | digit | pos | digit | pos | digit | pos | digit |
|-----|---

```json
{"n":72,"c":[13,6,7]}
```

## Counting digits in s

**s = `476423909362133322888007718815186500190932742072030926205404904190112781`**

### Length
Counting all characters: **n = 72**

### Digit-by-digit scan for 0, 7, 9

| Digi

```json
{"n":72,"c":[13,6,7]}
```

**Verification breakdown:**

| Digit | Positions (1-indexed) | Count |
|---|---|---|
| **0** | 8,22,23,35,36,39,46,49,51,56,59,62,66 | **13** |
| **7** | 2,24,25,43,

```
s = "476423909362133322888007718815186500190932742072030926205404904190112781"
```

**Counting in groups of 10:**

| Group | Digits | 0s | 7s | 9s |
|---|---|---|---|---|
| 4764239093 | 10 | 1 | 1

总计

总输入(含缓存) 93,245 0 93,267 93,237 91,335

总输出 1,851 1,094 1,284 1,210 2,349

总缓存创建

12,965

5m: 0 / 1h: 12,965

12,989

5m: 0 / 1h: 12,989

12,959

5m: 0 / 1h: 12,959

29,218

5m: 0 / 1h: 29,218

总缓存读取 80,101 0 80,099 80,099 61,992

输出速度 97.8 tok/s 73.1 tok/s 85.5 tok/s 80.7 tok/s 84.2 tok/s

总执行时间 39.2s 33.9s 43.0s 36.2s 53.3s

总墙钟时间

13m 41.1s

含步骤间等待

9m 13.6s

含步骤间等待

12m 40.9s

含步骤间等待

15m 3.9s

含步骤间等待

17m 3.0s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.1010

Anthropic: Claude Sonnet 4.6

>=$0.0164

Anthropic: Claude Sonnet 4.6 (成本不完整)

$0.0925

Anthropic: Claude Sonnet 4.6

$0.0913

Anthropic: Claude Sonnet 4.6

$0.1638

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

ID 格式 msg_01PkwkBX4Hkc92MajpTdksXf msg_8d7b81e1f4cc4225b04d2d385e467027 msg_01AbUhX5hYvXVyR7skwmGQ3b msg_0171a94nhNCEDUzp3wYi7QUX msg_01PvJyjBinzRP4GhoufTE4fB

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
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_011Cccu3jrenUWWWwsrm3ChR
R-Kiro	server: Photon-Edge x-amzn-requestid: d5d3f971-f28e-4aec-8cd8-4eafc79d324e	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	1782979853248-9afkcop0iue4d d5d3f971-f28e-4aec-8cd8-4eafc79d324e
O-Max	via: 1.1 Caddy cf-ray: a14c3250dc85dc54-FRA server: cloudflare x-new-api-version: v0.12.10 x-oneapi-request-id: 202607020814229360618228268d9d6p4zDDjMz	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-new-api-versionhdr: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	a14c3250dc85dc54-FRA
cc-Zeus	via: 1.1 Caddy cf-ray: a14c35b4fd2c78ef-LAX server: cloudflare x-oneapi-request-id: 20260702161641775240517EKiuzXBB	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	a14c35b4fd2c78ef-LAX
M-Max	cf-ray: a14c38652b43590c-AMS server: cloudflare x-new-api-version: v0.0.0 x-oneapi-request-id: 202607020818319174839448268d9d6AUxPSsQk	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-idhdr:server:cloudflarehdr:cf-raybody:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	a14c38652b43590c-AMS

协议指纹评分（机器自动）

总分 / 100 100 53 98 99 88

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 9.7 /10 ×20

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

延迟基线 10.0 /10 ×15 10.0 /10 ×15 9.3 /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 5.0 /10 ×14 10.0 /10 ×14 10.0 /10 ×14 10.0 /10 ×14

sliding 5m 10.0 /10 ×13 5.0 /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 10.0 /10 ×8 未计入 10.0 /10 ×8 10.0 /10 ×8 5.0 /10 ×8

身份(结构) 10.0 /10 ×7 10.0 /10 ×7 10.0 /10 ×7 10.0 /10 ×7 6.7 /10 ×7

身份(自由) 10.0 /10 ×7 0.0 /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

知识截止 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 5.0 /10 ×5

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

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

Req-ID 透传 10.0 /10 ×2 0.0 /10 ×2 0.0 /10 ×2 0.0 /10 ×2 0.0 /10 ×2

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

与基线相似度基准（参考） 53 v3.26.0 98 v3.26.0 99 v3.26.0 88 v3.26.0

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 5492 ms 4931 ms 5857 ms 4724 ms 7627 ms

输出速度 97.8 tok/s 73.1 tok/s 85.5 tok/s 80.7 tok/s 84.2 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 ⓘ

$0.1010

>=$0.0164 (成本不完整)

$0.0925

$0.0913

$0.1638

缓存读取占比 ⓘ 86% - 86% 86% 68%

揭晓

通道

官方基准

服务商 SSSAiCode ↗

通道R-Kiro

服务商 sucui ↗

通道O-Max

服务商 PackyCode ↗

通道cc-Zeus

服务商 AIMZ ↗

通道M-Max

返回列表