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

2026/06/23 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

ClaudeAPI

O-Bdrk

官方基准

AnyAIGC

M-

TopRouterCN

O-Max

0-0

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 10:03:02 ~ 10:03:12 (9.0s) 10:03:08 ~ 10:03:13 (4.4s) 10:03:00 ~ 10:03:05 (4.1s) 10:03:04 ~ 10:03:14 (5.0s) 10:03:06 ~ 10:03:16 (5.6s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 5,526 ms 2,847 ms 2,565 ms 3,927 ms 3,909 ms

首文本(SSE delta) 5,536 ms 2,848 ms 2,567 ms 3,934 ms 3,910 ms

总耗时 5,943 ms 2,959 ms 2,875 ms 4,008 ms 4,121 ms

输入 5,934 3 8,598 1,467 3

输出 5 5 2 5 5

缓存创建

25,697

5m: 25,697 / 1h: 0

32,318

5m: 0 / 1h: 32,318

30,854

5m: 0 / 1h: 30,854

32,806

5m: 32,806 / 1h: 0

缓存读取 0 0 0 0 0

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 10:06:20 ~ 10:06:30 (5.9s) 10:04:40 ~ 10:04:45 (4.5s) 10:04:10 ~ 10:04:20 (6.6s) 10:06:58 ~ 10:07:03 (4.8s) 10:06:22 ~ 10:06:32 (6.2s)

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) 4,326 ms 2,931 ms 5,290 ms 3,866 ms 4,164 ms

首文本(SSE delta) 4,327 ms 2,932 ms 5,291 ms 3,867 ms 4,165 ms

总耗时 4,415 ms 3,411 ms 5,292 ms 3,954 ms 4,770 ms

输入 3 3 8,799 3 3

输出 22 22 21 22 22

缓存创建

6,122

5m: 6,122 / 1h: 0

191

5m: 0 / 1h: 191

1,655

5m: 0 / 1h: 1,655

32,996

5m: 32,996 / 1h: 0

缓存读取 25,697 32,318 0 30,854 0

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 10:10:15 ~ 10:10:25 (5.6s) 10:08:26 ~ 10:08:31 (3.3s) 10:06:51 ~ 10:07:01 (6.3s) 10:09:20 ~ 10:09:30 (6.1s) 10:09:42 ~ 10:09:52 (5.5s)

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) 4,137 ms 1,975 ms 4,914 ms 5,282 ms 3,712 ms

首文本(SSE delta) 4,359 ms 2,123 ms 4,959 ms 5,283 ms 3,809 ms

总耗时 4,388 ms 2,287 ms 5,007 ms 5,351 ms 3,890 ms

输入 3 3 8,852 3 3

输出 8 8 5 8 8

缓存创建

5m: 55 / 1h: 0

5m: 0 / 1h: 57

5m: 0 / 1h: 55

7,691

5m: 7,691 / 1h: 0

缓存读取 31,819 32,509 0 32,509 25,362

响应

2025-08

2025-08

2024-05

2025-08

2025-08

步骤 4: identity_free text

时间 10:12:07 ~ 10:12:17 (8.7s) 10:09:53 ~ 10:09:58 (4.2s) 10:09:26 ~ 10:09:36 (6.3s) 10:10:38 ~ 10:10:43 (3.9s) 10:13:53 ~ 10:14:03 (5.9s)

提示词

查看本步 Prompt

中文做个自我介绍：你是什么产品或助手？谁开发的？50 字内。

HTTP 头(TTFB) 6,670 ms 2,311 ms 3,399 ms 3,066 ms 3,441 ms

首文本(SSE delta) 6,672 ms 2,313 ms 3,401 ms 3,067 ms 3,442 ms

总耗时 7,287 ms 3,179 ms 3,643 ms 3,163 ms 4,186 ms

输入 3 3 8,889 3 3

输出 53 54 14 24 43

缓存创建

5m: 43 / 1h: 0

5m: 0 / 1h: 43

5m: 0 / 1h: 35

7,732

5m: 7,732 / 1h: 0

缓存读取 31,874 32,566 0 32,564 25,362

输出速度 86.2 tok/s 62.4 tok/s - - 57.8 tok/s

响应

我是 Claude，由 Anthropic 开发的 AI 助手。我能帮助回答问题、分析信息、编写代码和完成各种文本任务。

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、分析等任务。

我叫Claude，属于Claude产品，由Anthropic公司开发。

我是 Claude，由 Anthropic 开发的 AI 助手。

我叫 Claude，是由 Anthropic 公司开发的 AI 助手，当前版本为 Claude Sonnet 4.6。

步骤 5: knowledge_recall text

时间 10:13:55 ~ 10:14:05 (6.8s) 10:13:29 ~ 10:13:39 (4.9s) 10:13:09 ~ 10:13:14 (4.6s) 10:13:27 ~ 10:13:37 (5.2s) 10:17:18 ~ 10:17:28 (6.3s)

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) 2,868 ms 1,914 ms 2,078 ms 4,283 ms 2,798 ms

首文本(SSE delta) 2,869 ms 1,915 ms 2,079 ms 4,288 ms 2,799 ms

总耗时 5,309 ms 3,785 ms 3,006 ms 4,337 ms 4,755 ms

输入 3 3 9,498 3 3

输出 205 204 154 205 205

缓存创建

8,075

5m: 8,075 / 1h: 0

8,052

5m: 0 / 1h: 8,052

8,012

5m: 0 / 1h: 8,012

8,458

5m: 8,458 / 1h: 0

缓存读取 0 0 0 0 0

输出速度 84.0 tok/s 109.1 tok/s burst burst 104.8 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":"unknown","oscars2024_best_picture":"unknown","ucl_2024":"unknown","nba_2024":"unknown","copa_america_2024":"unknown","euro2024":"unknown","nobel_peace_2024":"unknown","wo

{"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 10:15:54 ~ 10:16:14 (17.0s) 10:16:21 ~ 10:16:41 (15.9s) 10:15:17 ~ 10:15:22 (3.1s) 10:17:30 ~ 10:17:45 (10.6s) 10:19:02 ~ 10:19:22 (16.4s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=544372184112359560121282154341188258051121168760727907290507068404181356

HTTP 头(TTFB) 2,351 ms 2,126 ms 1,552 ms 6,556 ms 2,467 ms

首文本(SSE delta) 2,352 ms 2,128 ms 1,553 ms 6,557 ms 2,468 ms

总耗时 15,559 ms 14,957 ms 1,554 ms 9,830 ms 13,679 ms

输入 191 164 166 164 581

输出 1,057 1,106 18 800 979

缓存创建

5m: 0 / 1h: 0

缓存读取 0 0 0 0 0

输出速度 80.0 tok/s 86.2 tok/s - burst 87.3 tok/s

响应

{"n":72,"c":[8,6,3]}

**Counting the string:**

`s = 544372184112359560121282154341188258051121168760727907290507068404181356`

**Length:**
Splitting into groups of 10:
`5443721841 | 1235956012 | 1282154341 | 1882580511 |

```json
{"n":66,"c":[6,5,5]}
```

```json
{"n":72,"c":[8,6,3]}
```

```json
{"n":72,"c":[8,6,3]}
```

总计

总输入(含缓存) 135,519 138,233 44,802 138,181 141,003

总输出 1,350 1,399 214 1,064 1,262

总缓存创建

39,992

5m: 39,992 / 1h: 0

40,661

5m: 0 / 1h: 40,661

40,611

5m: 0 / 1h: 40,611

89,683

5m: 89,683 / 1h: 0

总缓存读取 89,390 97,393 0 95,927 50,724

输出速度 80.9 tok/s 87.6 tok/s burst burst 88.2 tok/s

总执行时间 53.1s 37.3s 31.0s 35.5s 45.9s

总墙钟时间

13m 12.0s

含步骤间等待

13m 33.0s

含步骤间等待

12m 22.0s

含步骤间等待

14m 41.0s

含步骤间等待

16m 16.0s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.1592

Anthropic: Claude Sonnet 4.6

$0.2032

Anthropic: Claude Sonnet 4.6

$0.1376

Anthropic: Claude Sonnet 4.6

$0.2020

Anthropic: Claude Sonnet 4.6

$0.2461

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%) unknown (0%) direct (100%) direct (100%)

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

CDN - - - - -

ID 格式 msg_01XxAsTsVzEAHY78RbWD3JB4 msg_019HtdrLKonMvori18nmVvuL 自定义 msg_0118h5S2ErYvu8HqyyUCGFJV msg_01T7a22tGCjDe5G6zYztEHSp

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-Bdrk	x-new-api-version: v1.0.0-rc.11 x-oneapi-request-id: 202606230215559562151248268d9d6bEnyuRkw	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_011CcKPPjmCMZZ1BdXsGkojx
M-	server: nginx	id_format: non_standard has_service_tier: false has_inference_geo: false model_has_date_suffix: false usage_has_cache_fields: true	body:id:non_standardbody:usage.cache_fields	—
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-tlfecbjbfy-069928
O-Max	via: 1.1 Caddy server: uvicorn, nginx/1.24.0 (Ubuntu)	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:viabody:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	e490631b41d1b5555282fb0c3527ae54

协议指纹评分（机器自动）

总分 / 100 87 100 37 95 74

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 0.4 /10 ×20

缓存 TTL 一致 0.0 /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 10.0 /10 ×14

sliding 5m 10.0 /10 ×13 10.0 /10 ×13 5.0 /10 ×13 10.0 /10 ×13 7.7 /10 ×13

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

系统提示纯净 5.0 /10 ×8 10.0 /10 ×8 0.0 /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 10.0 /10 ×7

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

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

延迟基线 9.6 /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 3.8 /10 ×5 0.0 /10 ×5 10.0 /10 ×5

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 87 v3.24.0 基准（参考） 37 v3.24.0 95 v3.24.0 74 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 7150 ms 5096 ms 3563 ms 5107 ms 5900 ms

输出速度 80.9 tok/s 87.6 tok/s burst burst 88.2 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 ⓘ

$0.1592

$0.2032

$0.1376

$0.2020

$0.2461

缓存读取占比 ⓘ 66% 70% 0% 69% 36%

揭晓

通道

服务商 ClaudeAPI ↗

通道O-Bdrk

官方基准

服务商 AnyAIGC ↗

通道M-

服务商 TopRouterCN ↗

通道O-Max

服务商 0-0 ↗

通道O-Max

返回列表