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

2026/06/25 运行日期

claude-haiku-4-5-20251001 官方基准

4 通道分数 < 70 与基线相似度

已揭晓揭晓进度

步骤 / 指标

SSSAiCode

O-Max-L2

官方基准

SSSAiCode

R-Kiro

SSSAiCode

O-Max

AnyAIGC

M-

基本信息

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

模型 claude-haiku-4-5-20251001 claude-haiku-4-5-20251001 claude-haiku-4-5-20251001 claude-haiku-4-5-20251001 claude-haiku-4-5-20251001

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 08:47:01 ~ 08:47:06 (4.7s) 08:47:05 ~ 08:47:10 (3.2s) 08:46:59 ~ 08:47:04 (4.8s) 08:47:03 ~ 08:47:08 (3.8s) 08:46:57 ~ 08:47:07 (5.1s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 2,503 ms 1,619 ms 2,902 ms 2,081 ms 3,654 ms

首文本(SSE delta) 2,917 ms 1,621 ms 2,927 ms 2,464 ms 3,656 ms

总耗时 3,059 ms 1,960 ms 3,292 ms 2,603 ms 3,675 ms

输入 10 10 0 10 8,600

输出 54 58 53 52 2

缓存创建

32,860

5m: 0 / 1h: 32,860

33,122

5m: 0 / 1h: 33,122

32,790

5m: 0 / 1h: 32,790

缓存读取 0 0 0 0 0

输出速度 - 171.1 tok/s 145.2 tok/s - -

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 08:48:23 ~ 08:48:39 (10.4s) 08:49:29 ~ 08:49:34 (3.9s) 08:50:10 ~ 08:50:15 (4.5s) 08:50:02 ~ 08:50:12 (5.9s) 08:50:27 ~ 08:50:32 (4.4s)

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) 1,613 ms 1,413 ms 2,189 ms 1,601 ms 2,917 ms

首文本(SSE delta) 9,169 ms 1,415 ms 2,215 ms 4,939 ms 2,918 ms

总耗时 9,454 ms 2,969 ms 3,787 ms 5,120 ms 3,045 ms

输入 10 10 0 10 8,810

输出 799 167 202 388 25

缓存创建

258

5m: 0 / 1h: 258

262

5m: 0 / 1h: 262

256

5m: 0 / 1h: 256

缓存读取 32,860 33,122 0 32,790 0

输出速度 burst 107.5 tok/s 128.5 tok/s - -

响应

vendor: anthropic
brand: claude
model: claude-opus-4.5

vendor: anthropic
brand: claude
model: claude-haiku-4-5-20251001

vendor: anthropic
brand: claude
model: claude-haiku-4-5-20251001

vendor: anthropic
brand: claude
model: claude-3-5-sonnet-20241022

vendor: anthropic  
brand: claude  
model: claude-haiku-4-5-20251001

步骤 3: cutoff text

时间 08:51:35 ~ 08:51:40 (4.0s) 08:51:37 ~ 08:51:42 (3.5s) 08:53:23 ~ 08:53:28 (4.0s) 08:51:40 ~ 08:51:45 (4.1s) 08:53:56 ~ 08:54:01 (4.1s)

提示词

查看本步 Prompt

Knowledge cutoff date? Reply with YYYY-MM only.

HTTP 头(TTFB) 2,119 ms 1,379 ms 2,620 ms 2,430 ms 2,697 ms

首文本(SSE delta) 3,045 ms 1,382 ms 2,648 ms 2,984 ms 2,700 ms

总耗时 3,199 ms 2,249 ms 3,176 ms 3,224 ms 2,862 ms

输入 10 10 0 10 8,867

输出 58 66 72 77 5

缓存创建

822

5m: 0 / 1h: 822

209

5m: 0 / 1h: 209

428

5m: 0 / 1h: 428

缓存读取 33,118 33,384 0 33,046 0

输出速度 - 76.1 tok/s 136.4 tok/s - -

响应

2024-04

2025-02

2025-02

2024-04

2025-05

步骤 4: identity_free text

时间 08:55:09 ~ 08:55:14 (4.7s) 08:53:04 ~ 08:53:09 (4.9s) 08:57:13 ~ 08:57:18 (3.6s) 08:54:28 ~ 08:54:43 (10.6s) 08:56:16 ~ 08:56:21 (4.2s)

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) 2,369 ms 1,414 ms 2,091 ms 1,966 ms 2,759 ms

首文本(SSE delta) 3,665 ms 1,416 ms 2,114 ms 7,645 ms 2,760 ms

总耗时 3,952 ms 3,850 ms 2,789 ms 7,792 ms 2,993 ms

输入 10 10 0 10 8,909

输出 191 417 87 819 27

缓存创建

5m: 0 / 1h: 92

100

5m: 0 / 1h: 100

115

5m: 0 / 1h: 115

缓存读取 33,940 33,593 0 33,474 0

输出速度 - 171.3 tok/s 128.9 tok/s burst -

响应

我是Claude，由Anthropic公司开发的AI助手。我能帮助用户进行文本分析、编程、研究、创意写作等各种任务。

我是Claude，由Anthropic开发的AI助手。我能帮助用户完成软件工程任务、代码编写、问题解答等工作。

我叫Claude，是由Anthropic公司开发的AI助手。我是Claude系列模型之一，本次使用的是Claude Haiku 4.5版本。

我是Claude，由Anthropic开发的AI助手。可帮助编程、分析、写作等各类任务。

我是由Anthropic开发的AI助手，名为Claude，帮助用户解决软件工程、学习和信息查询等任务。

步骤 5: knowledge_recall text

时间 08:59:01 ~ 08:59:16 (14.7s) 08:55:35 ~ 08:56:00 (20.0s) 09:01:11 ~ 09:01:26 (12.4s) 08:57:50 ~ 08:58:10 (17.4s) 08:58:41 ~ 08:58:46 (4.7s)

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) 1,562 ms 2,296 ms 2,331 ms 1,562 ms 2,417 ms

首文本(SSE delta) 13,375 ms 2,298 ms 2,357 ms 16,278 ms 2,419 ms

总耗时 13,901 ms 18,969 ms 11,431 ms 16,598 ms 3,429 ms

输入 10 10 0 10 9,541

输出 2,021 2,334 1,671 2,484 154

缓存创建

8,841

5m: 0 / 1h: 8,841

8,728

5m: 0 / 1h: 8,728

9,026

5m: 0 / 1h: 9,026

缓存读取 0 0 0 0 0

输出速度 burst 140.0 tok/s 184.2 tok/s burst 152.5 tok/s

响应

{"facts":{"super_bowl_lviii":"Kansas City Chiefs","oscars2024_best_picture":"Oppenheimer","ucl_2024":"UNKNOWN","nba_2024":"UNKNOWN","copa_america_2024":"UNKNOWN","euro2024":"UNKNOWN","nobel_peace_2024

```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

```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

```json
{"facts":{"super_bowl_lviii":"Kansas City Chiefs","oscars2024_best_picture":"Oppenheimer","ucl_2024":"UNKNOWN","nba_2024":"UNKNOWN","copa_america_2024":"UNKNOWN","euro2024":"UNKNOWN","nobel_pe

{"facts":{"super_bowl_lviii":"UNKNOWN","oscars2024_best_picture":"UNKNOWN","ucl_2024":"UNKNOWN","nba_2024":"UNKNOWN","copa_america_2024":"UNKNOWN","euro2024":"UNKNOWN","nobel_peace_2024":"UNKNOWN","wo

步骤 6: digit_count text

时间 09:02:52 ~ 09:03:17 (20.3s) 08:59:58 ~ 09:01:48 (108.5s) 09:03:48 ~ 09:04:03 (10.0s) 09:00:02 ~ 09:00:22 (18.5s) 09:01:49 ~ 09:01:54 (4.3s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=087316550622682650282506079774652448318944143798470075939268624159585751

HTTP 头(TTFB) 7,434 ms 1,236 ms 1,965 ms 1,464 ms 2,239 ms

首文本(SSE delta) 7,617 ms 1,238 ms 1,988 ms 1,465 ms 2,241 ms

总耗时 19,608 ms 107,580 ms 9,082 ms 17,555 ms 3,135 ms

输入 139 193 0 139 166

输出 2,414 9,525 1,180 2,752 14

缓存创建

5m: 0 / 1h: 0

缓存读取 0 0 0 0 0

输出速度 201.3 tok/s 89.6 tok/s 166.3 tok/s 171.0 tok/s -

响应

```json
{"n":72,"c":[7,8,6]}
```

```json
{"n":73,"c":[7,8,6]}
```

```json
{"n":72,"c":[7,8,6]}
```

```json
{"n":72,"c":[7,8,6]}
```

{"n":66,"c":[2,13,6]}

总计

总输入(含缓存) 142,980 142,763 0 142,114 44,893

总输出 5,537 12,567 3,265 6,572 227

总缓存创建

42,873

5m: 0 / 1h: 42,873

42,421

5m: 0 / 1h: 42,421

42,615

5m: 0 / 1h: 42,615

总缓存读取 99,918 100,099 0 99,310 0

输出速度 201.3 tok/s 98.0 tok/s 169.1 tok/s 171.0 tok/s 152.5 tok/s

总执行时间 58.8s 2m 24.1s 39.2s 1m 0.3s 26.8s

总墙钟时间

16m 16.0s

含步骤间等待

14m 43.3s

含步骤间等待

17m 4.0s

含步骤间等待

13m 19.1s

含步骤间等待

14m 57.0s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.0915

Anthropic: Claude Haiku 4.5

$0.1261

Anthropic: Claude Haiku 4.5

>=$0.0163

Anthropic: Claude Haiku 4.5 (成本不完整)

$0.0962

Anthropic: Claude Haiku 4.5

$0.0460

Anthropic: Claude Haiku 4.5

通道指纹

来源步骤 digit_count digit_count digit_count digit_count digit_count

平台 direct (100%) direct (100%) direct (75%) direct (100%) direct (50%)

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

CDN - - - - -

ID 格式 msg_01H8ziBkeAf3Xa6WsU8XFxqx msg_01Vx4bCcWpMPyGNzjCEWKvQZ msg_7acbc4dfe0f94d8a8dd3c0044ad27984 msg_01QdVxgzspHLzL3zDcRdGK8C 自定义

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-L2	cf-ray: a1100ca468f02acf-LAX server: Photon-Edge set-cookie: _cfuvid=3v41MxpL5EATNxeLukIOJ8OmpSkBefPd6a132JPZ2Xc-1782349374.145775-1.0.1.1-Dv7EoC5Y_yO5dngFBYB2RflpgowSsi1Bi0IOMjUrExs; 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: true usage_has_cache_fields: true	body:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:model:date_suffixbody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	1782349373547-yb4a9a57wug4x req_011CcP5QbrtaaMKRD6koVC5P a1100ca468f02acf-LAX
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: true usage_has_cache_fields: true	body:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:model:date_suffixbody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	req_011CcP5Bm6A9XDGFVeVdUhBf
R-Kiro	server: Photon-Edge x-amzn-requestid: 646d0015-fb3b-4a0d-acdf-65db2501d121	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	1782349429732-iwo23zmlr41ii 646d0015-fb3b-4a0d-acdf-65db2501d121
O-Max	cf-ray: a110087fd86afc10-LAX server: Photon-Edge set-cookie: _cfuvid=YUIkR.W8PuXCcuE.zdigUi39G4iws1LPMnNG7Uax.s8-1782349204.4532146-1.0.1.1-u6phV2HFFpNCTjjkyOG6afzcr4qfjlxHXXPBVxla5.4; 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: true usage_has_cache_fields: true	body:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:model:date_suffixbody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	1782349203867-weulxobj0mps6 req_011CcP5C6NXMhLRMxRjVRLML a110087fd86afc10-LAX
M-	server: nginx	id_format: non_standard has_service_tier: false has_inference_geo: false model_has_date_suffix: true usage_has_cache_fields: true	body:id:non_standardbody:model:date_suffixbody:usage.cache_fields	—

协议指纹评分（机器自动）

总分 / 100 82 96 53 84 35

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 0.0 /10 ×20

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

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

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

系统提示纯净 10.0 /10 ×8 10.0 /10 ×8 未计入 5.0 /10 ×8 0.0 /10 ×8

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

知识截止 0.0 /10 ×7 10.0 /10 ×7 10.0 /10 ×7 0.0 /10 ×7 5.0 /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 0.0 /10 ×6 10.0 /10 ×6 0.0 /10 ×6

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

Req-ID 透传 10.0 /10 ×4 10.0 /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

流式投递未计入未计入未计入未计入未计入

综合结论

与基线相似度 82 v3.24.1 基准（参考） 53 v3.24.1 84 v3.24.1 35 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 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 Thinking 体量 thinking_volume_match
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 8862 ms 22930 ms 5593 ms 8815 ms 3190 ms

输出速度 201.3 tok/s 98.0 tok/s 169.1 tok/s 171.0 tok/s 152.5 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 ⓘ

$0.0915

$0.1261

>=$0.0163 (成本不完整)

$0.0962

$0.0460

缓存读取占比 ⓘ 70% 70% - 70% 0%

揭晓

通道

服务商 SSSAiCode ↗

通道O-Max-L2

官方基准

服务商 SSSAiCode ↗

通道R-Kiro

服务商 SSSAiCode ↗

通道O-Max

服务商 AnyAIGC ↗

通道M-

返回列表