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-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

AnyAIGC

M-

YunWu

O-Api

SudoCode

M-Max

FoxCode

官方基准

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 09:55:40 ~ 09:55:45 (4.3s) 09:55:44 ~ 09:55:54 (5.0s) 09:55:46 ~ 09:56:11 (23.4s) 09:55:42 ~ 09:55:47 (4.0s) 09:55:48 ~ 09:55:53 (4.2s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 2,966 ms 3,164 ms 21,691 ms 2,497 ms 2,715 ms

首文本(SSE delta) 2,968 ms 3,172 ms 21,692 ms 2,499 ms 2,717 ms

总耗时 2,991 ms 3,173 ms 21,730 ms 2,709 ms 2,829 ms

输入 8,598 8,598 3 3 3

输出 2 2 5 5 5

缓存创建

32,318

5m: 0 / 1h: 32,318

32,318

5m: 32,318 / 1h: 0

32,318

5m: 0 / 1h: 32,318

缓存读取 0 0 0 0 0

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 09:57:11 ~ 09:57:16 (4.4s) 09:57:42 ~ 09:57:47 (3.6s) 10:00:03 ~ 10:00:28 (22.7s) 09:59:42 ~ 09:59:47 (3.6s) 09:59:39 ~ 09:59:44 (3.6s)

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。自报身份，输出严格遵守以下 3 行格式（不允许 markdown、代码块、额外文本）：
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) 3,079 ms 2,401 ms 21,567 ms 2,427 ms 1,852 ms

首文本(SSE delta) 3,080 ms 2,402 ms 21,638 ms 2,448 ms 1,857 ms

总耗时 3,080 ms 2,406 ms 21,647 ms 2,703 ms 2,585 ms

输入 8,797 8,797 3 3 3

输出 21 21 22 22 22

缓存创建

192

5m: 0 / 1h: 192

192

5m: 192 / 1h: 0

192

5m: 0 / 1h: 192

缓存读取 0 0 32,318 32,318 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

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 10:00:32 ~ 10:00:37 (3.4s) 10:01:15 ~ 10:01:20 (3.8s) 10:01:37 ~ 10:02:02 (22.5s) 10:01:27 ~ 10:01:32 (3.3s) 10:00:59 ~ 10:01:04 (3.3s)

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) 2,133 ms 2,595 ms 21,369 ms 2,089 ms 2,112 ms

首文本(SSE delta) 2,167 ms 2,596 ms 21,371 ms 2,090 ms 2,242 ms

总耗时 2,222 ms 2,596 ms 21,400 ms 2,325 ms 2,316 ms

输入 8,854 8,851 3 3 3

输出 5 5 8 8 8

缓存创建

5m: 0 / 1h: 38

5m: 38 / 1h: 0

5m: 0 / 1h: 38

缓存读取 0 0 32,510 32,510 32,510

响应

2025-05

2025-05

2025-08

2025-08

2025-08

步骤 4: identity_free text

时间 10:03:40 ~ 10:03:50 (5.7s) 10:05:18 ~ 10:05:23 (4.9s) 10:04:08 ~ 10:04:28 (19.7s) 10:04:17 ~ 10:04:22 (4.5s) 10:02:33 ~ 10:02:38 (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) 4,543 ms 3,583 ms 17,501 ms 2,310 ms 2,469 ms

首文本(SSE delta) 4,544 ms 3,584 ms 17,502 ms 2,311 ms 2,924 ms

总耗时 4,544 ms 3,584 ms 18,694 ms 3,398 ms 3,257 ms

输入 8,892 8,892 3 3 3

输出 23 25 41 57 51

缓存创建

5m: 0 / 1h: 35

5m: 35 / 1h: 0

5m: 0 / 1h: 41

缓存读取 0 0 32,548 32,548 32,548

输出速度 - - 34.4 tok/s 52.4 tok/s -

响应

我是由Anthropic打造的人工智能助手Claude，专注于智能对话和软件工程辅助。

我是由Anthropic开发的Claude智能助手，专注于提供高效、安全的编程和技术支持。

我是Claude，由Anthropic公司开发的AI助手。Anthropic是一家专注于AI安全研究的美国公司。

我是 Claude，由 Anthropic 公司开发的 AI 助手。我基于大型语言模型构建，旨在提供安全、有益的对话和任务协助。

我叫 Claude，是由 Anthropic 开发的 AI 助手。当前版本为 claude-sonnet-4-6，属于 Claude 产品系列。

步骤 5: knowledge_recall text

时间 10:05:45 ~ 10:05:55 (5.7s) 10:07:20 ~ 10:07:25 (4.4s) 10:08:00 ~ 10:08:15 (12.6s) 10:08:06 ~ 10:08:16 (4.9s) 10:04:48 ~ 10:04:53 (4.3s)

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) 2,538 ms 2,128 ms 8,299 ms 1,679 ms 1,402 ms

首文本(SSE delta) 2,539 ms 2,129 ms 8,300 ms 1,680 ms 1,403 ms

总耗时 4,500 ms 3,342 ms 11,511 ms 3,505 ms 3,297 ms

输入 9,517 9,520 3 3 3

输出 154 154 205 205 205

缓存创建

8,013

5m: 0 / 1h: 8,013

8,029

5m: 8,029 / 1h: 0

8,029

5m: 0 / 1h: 8,029

缓存读取 0 0 0 0 0

输出速度 78.5 tok/s 127.0 tok/s 63.8 tok/s 112.3 tok/s 108.2 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":"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

{"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 10:09:13 ~ 10:09:33 (16.0s) 10:11:06 ~ 10:11:11 (4.1s) 10:09:40 ~ 10:10:00 (15.3s) 10:11:58 ~ 10:12:18 (15.4s) 10:08:39 ~ 10:08:54 (14.2s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=329306353560075793520572451674997793294116351181830208404250869479328678

HTTP 头(TTFB) 14,068 ms 2,430 ms 1,682 ms 1,031 ms 1,431 ms

首文本(SSE delta) 14,070 ms 2,431 ms 1,711 ms 1,033 ms 1,433 ms

总耗时 14,859 ms 2,856 ms 14,151 ms 14,392 ms 13,303 ms

输入 166 166 164 164 164

输出 18 18 928 1,177 1,214

缓存创建

5m: 0 / 1h: 0

缓存读取 0 0 0 0 0

输出速度 - - 74.6 tok/s 88.1 tok/s 102.3 tok/s

响应

```json
{"n":66,"c":[6,5,7]}
```

```json
{"n":66,"c":[4,5,5]}
```

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

**Verification:**

`s = 329306353560075793520572451674997793294116351181830208404250869479328678`

**Length:** Grouped in 10s → 70 + 2 = **72**

| Digit | Positions (

## Counting digits in s

**s** = `329306353560075793520572451674997793294116351181830208404250869479328678`

### Step 1 — Total length

Splitting into groups of 10:
```
3293063535  (10)
6007579352  (1

```
s = "329306353560075793520572451674997793294116351181830208404250869479328678"
```

**Counting character by character (indexed 0–71):**

| Digit | Positions |
|-------|-----------|
| **0** | 4, 11

总计

总输入(含缓存) 44,824 44,824 138,151 138,167 138,173

总输出 223 225 1,209 1,474 1,505

总缓存创建

40,596

5m: 0 / 1h: 40,596

40,612

5m: 40,612 / 1h: 0

40,618

5m: 0 / 1h: 40,618

总缓存读取 0 0 97,376 97,376 97,376

输出速度 78.5 tok/s 127.0 tok/s 69.7 tok/s 88.4 tok/s 103.1 tok/s

总执行时间 39.6s 25.8s 1m 56.3s 35.7s 34.0s

总墙钟时间

13m 53.0s

含步骤间等待

15m 26.9s

含步骤间等待

14m 14.1s

含步骤间等待

16m 36.0s

含步骤间等待

13m 5.9s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.1378

Anthropic: Claude Sonnet 4.6

$0.1378

Anthropic: Claude Sonnet 4.6

$0.2001

Anthropic: Claude Sonnet 4.6

$0.1470

Anthropic: Claude Sonnet 4.6

$0.2046

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

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

CDN - - - cloudflare -

ID 格式自定义自定义 msg_01CuV5yVCk9QarMPfjqueri8 msg_011dy2drwoD8k91bjwTnH9CY msg_01Qp3473FzQkds3RLZ1wUgn7

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
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-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-Max	server: nginx x-new-api-version: 0b8be5cf x-oneapi-request-id: 2026062510094296730274ZprDzFHI	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	—
cc	cf-ray: a11071da3d00734b-NRT server: cloudflare set-cookie: _cfuvid=xCiTSBMmLWYG262SNckuJMmY4J5P_PJQYi37QpT4h4s-1782353519.911304-1.0.1.1-P67ru1QPfRXMeLqxZPq_66gTjZDnCZuSLuWhT85apZ4; 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: false usage_has_cache_fields: true	hdr:server:cloudflarehdr:cf-raybody:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	req_011CcPAgCidaQYGaPgL9MdWC a11071da3d00734b-NRT
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_011CcPARXpfAKAW2CG3U8x59

协议指纹评分（机器自动）

总分 / 100 39 39 95 91 100

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 0.0 /10 ×20 0.0 /10 ×20 10.0 /10 ×20 10.0 /10 ×20 10.0 /10 ×20

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

缓存连续性 5.0 /10 ×14 5.0 /10 ×14 10.0 /10 ×14 10.0 /10 ×14 10.0 /10 ×14

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

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

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

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

service_tier 0.0 /10 ×6 0.0 /10 ×6 10.0 /10 ×6 10.0 /10 ×6 10.0 /10 ×6

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

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

流式投递 9.5 /10 ×5 9.6 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5

Req-ID 透传 0.0 /10 ×4 0.0 /10 ×4 0.0 /10 ×4 10.0 /10 ×4 10.0 /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 39 v3.24.1 39 v3.24.1 95 v3.24.1 91 v3.24.1 基准（参考）

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 5366 ms 2993 ms 18189 ms 4839 ms 4598 ms

输出速度 78.5 tok/s 127.0 tok/s 69.7 tok/s 88.4 tok/s 103.1 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 ⓘ

$0.1378

$0.2001

$0.1470

$0.2046

缓存读取占比 ⓘ 0% 0% 70% 70% 70%

揭晓

通道

服务商 AnyAIGC ↗

通道M-

服务商 YunWu ↗

通道O-Api

服务商 SudoCode ↗

通道M-Max

服务商 FoxCode ↗

通道cc

官方基准

返回列表