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。

3 通道 · 6 步骤

2026/07/02 运行日期

claude-sonnet-4-6 官方基准

2 任务失败执行不完整

已揭晓揭晓进度

⚠ 2 通道不可测 · 已从主对比表移除点击展开原因

TopRouterCN / O-Max
```
评测失败
```
DBAI / M-
```
评测失败
```

步骤 / 指标

OmniaKey

O-Max

ClaudeCN

O-Max

官方基准

基本信息

服务名 Claude Code Claude Code Claude Code

模型 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

偏差标红基于官方基准

步骤 1: ping text

时间 11:30:12 ~ 11:30:17 (4.7s) 11:30:14 ~ 11:30:19 (4.7s) 11:30:20 ~ 11:30:25 (3.8s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 3,578 ms 3,098 ms 2,708 ms

首文本(SSE delta) 3,579 ms 3,099 ms 2,710 ms

总耗时 3,687 ms 3,305 ms 2,798 ms

输入 3 3 3

输出 5 5 5

缓存创建

5,991

5m: 0 / 1h: 5,991

5,993

5m: 0 / 1h: 5,993

5m: 0 / 1h: 0

缓存读取 14,884 14,884 20,875

响应

pong

pong

pong

步骤 2: identity text

时间 11:31:28 ~ 11:31:34 (4.4s) 11:32:55 ~ 11:33:01 (4.2s) 11:33:42 ~ 11:33:48 (3.1s)

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。请按下列 3 行格式精确填空作答（label 不变，值小写）：
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) 2,927 ms 2,810 ms 1,651 ms

首文本(SSE delta) 2,928 ms 2,811 ms 1,652 ms

总耗时 3,145 ms 3,232 ms 2,125 ms

输入 3 3 3

输出 22 22 22

缓存创建

182

5m: 0 / 1h: 182

182

5m: 0 / 1h: 182

182

5m: 0 / 1h: 182

缓存读取 20,875 20,877 20,875

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

步骤 3: cutoff text

时间 11:33:36 ~ 11:33:46 (5.2s) 11:34:15 ~ 11:34:25 (5.1s) 11:35:19 ~ 11:35:24 (3.1s)

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) 3,595 ms 3,678 ms 1,855 ms

首文本(SSE delta) 3,763 ms 3,965 ms 2,022 ms

总耗时 4,098 ms 4,071 ms 2,077 ms

输入 3 3 3

输出 8 8 8

缓存创建

5m: 0 / 1h: 55

5m: 0 / 1h: 38

5m: 0 / 1h: 55

缓存读取 21,057 21,059 21,057

响应

2025-08

2025-08

2025-08

步骤 4: identity_free text

时间 11:35:50 ~ 11:36:00 (5.3s) 11:36:54 ~ 11:37:04 (6.3s) 11:38:59 ~ 11:39:04 (4.1s)

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) 2,821 ms 3,243 ms 2,250 ms

首文本(SSE delta) 2,822 ms 3,244 ms 2,252 ms

总耗时 4,197 ms 4,304 ms 3,115 ms

输入 3 3 3

输出 63 50 54

缓存创建

5m: 0 / 1h: 43

5m: 0 / 1h: 41

缓存读取 21,112 21,097 21,112

输出速度 45.8 tok/s 47.2 tok/s 62.6 tok/s

响应

我是 Claude，由 Anthropic 开发的 AI 助手。我擅长回答问题、分析代码、写作和推理，致力于提供安全、准确、有帮助的回复。

我是 Claude，由 Anthropic 开发的 AI 助手。我能帮助回答问题、分析代码、撰写文本等多种任务。

我叫 Claude，是由 Anthropic 开发的 AI 助手，当前使用的模型为 claude-sonnet-4-6，属于 Claude 产品系列。

步骤 5: knowledge_recall text

时间 11:38:43 ~ 11:38:53 (5.0s) 11:40:10 ~ 11:40:20 (9.2s) 11:42:54 ~ 11:42:59 (4.8s)

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) 2,191 ms 3,799 ms 2,099 ms

首文本(SSE delta) 2,192 ms 5,179 ms 2,100 ms

总耗时 4,009 ms 6,855 ms 3,831 ms

输入 3 3 3

输出 205 196 196

缓存创建

6,675

5m: 0 / 1h: 6,675

6,647

5m: 0 / 1h: 6,647

2,993

5m: 0 / 1h: 2,993

缓存读取 2,131 2,131 5,802

输出速度 112.8 tok/s 116.9 tok/s 113.2 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

步骤 6: digit_count text

时间 11:42:13 ~ 11:42:28 (11.6s) 11:42:35 ~ 11:43:05 (25.3s) 11:44:43 ~ 11:45:03 (17.6s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=506731648689441948187456059907105451377945928651828067156532966918599768

HTTP 头(TTFB) 1,668 ms 2,558 ms 1,317 ms

首文本(SSE delta) 1,669 ms 2,559 ms 1,318 ms

总耗时 10,446 ms 24,164 ms 16,665 ms

输入 164 164 164

输出 848 1,511 1,573

缓存创建

5m: 0 / 1h: 0

缓存读取 0 0 0

输出速度 96.6 tok/s 69.9 tok/s 102.5 tok/s

响应

```json
{"n":72,"c":[5,7,10]}
```

**Verification:**

String: `506731648689441948187456059907105451377945928651828067156532966918599768`

- **Length**: 72 characters
- **0s** (×5): positions 2, 25, 29

```
s = 506731648689441948187456059907105451377945928651828067156532966918599768
```

**Digit-by-digit breakdown (groups of 10):**

| Group | Digits | 0s | 7s | 9s |
|-------|--------|----|----|----|

## Counting digits in s

**s = `506731648689441948187456059907105451377945928651828067156532966918599768`**

### Length (n)

Splitting into groups of 10:
```
5067316486 | 8944194818 | 7456059907 | 105

总计

总输入(含缓存) 93,184 93,130 93,171

总输出 1,151 1,792 1,858

总缓存创建

12,946

5m: 0 / 1h: 12,946

12,903

5m: 0 / 1h: 12,903

3,271

5m: 0 / 1h: 3,271

总缓存读取 80,059 80,048 89,721

输出速度 93.2 tok/s 72.2 tok/s 101.6 tok/s

总执行时间 36.1s 54.8s 36.5s

总墙钟时间

12m 16.0s

含步骤间等待

12m 51.0s

含步骤间等待

14m 42.9s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.0904

Anthropic: Claude Sonnet 4.6

$0.0998

Anthropic: Claude Sonnet 4.6

$0.0676

Anthropic: Claude Sonnet 4.6

通道指纹

来源步骤 digit_count digit_count digit_count

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

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

CDN cloudflare cloudflare -

ID 格式 msg_01CLBASXjK8RBR99rutujH3D msg_01EA4GXNLYUvvH7pMB7n4miJ msg_01JyE68ccaoqAr4j9mkPLXEw

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

cf-ray: a14aa3af3f148a22-AMS
server: cloudflare
x-oneapi-request-id: 202607020342153938841768268d9d6y0IykBq6

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-idhdr:server:cloudflarehdr:cf-raybody:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body

a14aa3af3f148a22-AMS

O-Max

cf-ray: a14aa4391a8a0b88-AMS
server: cloudflare
x-oneapi-request-id: 20260702114237556067638ym5gMAo0

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-idhdr:server:cloudflarehdr:cf-raybody:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body

a14aa4391a8a0b88-AMS

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_011CccYQq1rUPwfFo5iEMrhD

协议指纹评分（机器自动）

总分 / 100 98 99 100

基线状态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 10.0 /10 ×20

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

延迟基线 9.5 /10 ×15 9.6 /10 ×15 10.0 /10 ×15

模型匹配 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

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

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

系统提示纯净 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

身份(自由) 10.0 /10 ×7 10.0 /10 ×7 10.0 /10 ×7

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

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

inference_geo 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

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

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

SDK 一致 10.0 /10 ×2 10.0 /10 ×2 10.0 /10 ×2

综合结论

与基线相似度 98 v3.26.0 99 v3.26.0 基准（参考）

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 4930 ms 7655 ms 5102 ms

输出速度 93.2 tok/s 72.2 tok/s 101.6 tok/s

完成步骤 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 ⓘ

$0.0904

$0.0998

$0.0676

缓存读取占比 ⓘ 86% 86% 96%

揭晓

通道

服务商 OmniaKey ↗

通道O-Max

服务商 ClaudeCN ↗

通道O-Max

官方基准

返回列表