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

2026/06/27 运行日期

claude-sonnet-4-6 官方基准

2 通道分数 < 70 另含 2 任务失败

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

DBAI / M-
```
评测失败
```
0-0 / O-Max
```
评测失败
```

步骤 / 指标

YunWu

O-Api

luminai

M-Kiro/Antigravity

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 v2.1.160 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 09:52:54 ~ 09:52:59 (3.9s) 09:52:52 ~ 09:53:02 (5.4s) 09:52:56 ~ 09:53:01 (4.2s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 2,631 ms 4,208 ms 2,976 ms

首文本(SSE delta) 2,632 ms 4,209 ms 2,977 ms

总耗时 2,658 ms 4,316 ms 3,113 ms

输入 8,598 2,309 3

输出 2 1 5

缓存创建

2,654

5m: 2,654 / 1h: 0

32,318

5m: 0 / 1h: 32,318

缓存读取 0 23,888 0

响应

pong

pong

pong

步骤 2: identity text

时间 09:54:46 ~ 09:54:56 (4.9s) 09:55:58 ~ 09:56:08 (6.7s) 09:54:19 ~ 09:54:24 (3.4s)

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) 3,198 ms 5,362 ms 1,835 ms

首文本(SSE delta) 3,199 ms 5,363 ms 1,836 ms

总耗时 3,627 ms 5,644 ms 2,395 ms

输入 8,788 2,323 3

输出 21 24 22

缓存创建

2,670

5m: 2,670 / 1h: 0

182

5m: 0 / 1h: 182

缓存读取 0 24,037 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

步骤 3: cutoff text

时间 09:57:02 ~ 09:57:12 (5.0s) 09:59:46 ~ 09:59:56 (5.6s) 09:57:13 ~ 09:57:18 (2.9s)

提示词

查看本步 Prompt

Knowledge cutoff date? Reply with YYYY-MM only.

HTTP 头(TTFB) 2,502 ms 4,278 ms 1,767 ms

首文本(SSE delta) 2,618 ms 4,415 ms 1,902 ms

总耗时 3,618 ms 4,476 ms 1,977 ms

输入 8,826 2,328 3

输出 5 4 8

缓存创建

2,676

5m: 2,676 / 1h: 0

5m: 0 / 1h: 38

缓存读取 0 24,086 32,500

响应

2025-05

2025-08

2025-08

步骤 4: identity_free text

时间 09:58:56 ~ 09:59:01 (4.4s) 10:03:31 ~ 10:03:41 (7.9s) 10:00:00 ~ 10:00:05 (4.2s)

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) 3,192 ms 4,879 ms 1,919 ms

首文本(SSE delta) 3,193 ms 5,094 ms 2,363 ms

总耗时 3,303 ms 5,385 ms 3,180 ms

输入 8,864 2,331 3

输出 25 25 65

缓存创建

2,679

5m: 2,679 / 1h: 0

5m: 0 / 1h: 43

缓存读取 0 24,120 32,538

输出速度 - - 79.6 tok/s

响应

我是Anthropic公司开发的Claude智能助手，一款专注安全与高效对话的AI语言模型。

我是 Claude Sonnet 4.6，由 Anthropic 打造的 AI 助手。

我是 Claude，由 Anthropic 开发的 AI 助手。我能帮助你完成写作、分析、编程、问答等多种任务，致力于提供安全、有帮助的回应。

步骤 5: knowledge_recall text

时间 10:01:08 ~ 10:01:13 (4.3s) 10:06:17 ~ 10:06:27 (5.9s) 10:01:14 ~ 10:01:19 (4.2s)

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) 2,073 ms 2,829 ms 1,422 ms

首文本(SSE delta) 2,074 ms 4,769 ms 1,423 ms

总耗时 3,251 ms 4,831 ms 3,180 ms

输入 9,492 597 3

输出 154 221 196

缓存创建

686

5m: 686 / 1h: 0

8,035

5m: 0 / 1h: 8,035

缓存读取 0 6,177 0

输出速度 130.8 tok/s burst 111.6 tok/s

响应

{"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 10:03:12 ~ 10:03:17 (3.0s) 10:09:59 ~ 10:10:04 (4.0s) 10:04:10 ~ 10:04:25 (14.0s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=550382843150583098984664680922749322169328352602557529797952847839131176

HTTP 头(TTFB) 1,411 ms 2,536 ms 1,113 ms

首文本(SSE delta) 1,702 ms 2,761 ms 1,114 ms

总耗时 1,757 ms 2,825 ms 12,988 ms

输入 166 189 164

输出 18 14 1,251

缓存创建

5m: 0 / 1h: 0

缓存读取 0 - 0

输出速度 - - 105.4 tok/s

响应

```json
{"n":69,"c":[4,5,4]}
```

{"n":71,"c":[2,3,5]}

```json
{"n":72,"c":[5,6,9]}
```

**Verification:**

String (split into groups of 10):
```
5503828431  (0s: 1)
5058309898  (0s: 2)
4664680922  (0s: 1)
7493221693  (0s: 0)
2835260255  (0s: 1)
752979795

总计

总输入(含缓存) 44,734 123,750 138,151

总输出 225 289 1,547

总缓存创建

11,365

5m: 11,365 / 1h: 0

40,616

5m: 0 / 1h: 40,616

总缓存读取 0 102,308 97,356

输出速度 130.8 tok/s burst 104.7 tok/s

总执行时间 25.6s 35.7s 32.9s

总墙钟时间

10m 23.0s

含步骤间等待

17m 11.9s

含步骤间等待

11m 28.9s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.1376

Anthropic: Claude Sonnet 4.6

$0.0919

Anthropic: Claude Sonnet 4.6

$0.2053

Anthropic: Claude Sonnet 4.6

通道指纹

来源步骤 digit_count digit_count digit_count

平台 unknown (0%) one-api (100%) direct (100%)

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

CDN - cloudflare -

ID 格式自定义 msg_01GpdtKM5QchsL3W7wZcdFfgGs msg_013wRifdufRwwhbLNsksUKam

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-Api

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

—

M-Kiro/Antigravity

cf-ray: a120e9b4d8f92409-AMS
server: cloudflare
x-oneapi-request-id: 202606270210015774101058268d9d6D2vMltM4

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

a120e9b4d8f92409-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_011CcSwhYZ1qs9tSUUShTuu4

协议指纹评分（机器自动）

总分 / 100 38 71 98

基线状态已对照基线已对照基线已对照基线

缓存命中比 0.0 /10 ×20 10.0 /10 ×20 8.4 /10 ×20

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

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

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

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

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

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

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

延迟基线 9.7 /10 ×5 5.1 /10 ×5 10.0 /10 ×5

流式投递 8.9 /10 ×5 0.0 /10 ×5 10.0 /10 ×5

Req-ID 透传 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

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 38 v3.25.0 71 v3.25.0 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 3036 ms 4580 ms 4472 ms

输出速度 130.8 tok/s burst 104.7 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 ⓘ

$0.1376

$0.0919

$0.2053

缓存读取占比 ⓘ 0% 83% 70%

揭晓

通道

服务商 YunWu ↗

通道O-Api

服务商 luminai ↗

通道M-Kiro/Antigravity

官方基准

返回列表