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

2026/06/30 运行日期

claude-sonnet-4-6 官方基准

1 任务失败执行不完整

已揭晓揭晓进度

步骤 / 指标

ByteCatCode

O-Max

SSSAiCode

O-Max-L2

官方基准

SudoCode

M-Max

SAIAi

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 14:41:36 ~ 14:41:46 (5.9s) 14:41:34 ~ 14:41:39 (3.4s) 14:41:40 ~ 14:41:45 (3.9s) 14:41:32 ~ 14:45:23 (228.3s) 14:41:39 ~ 14:41:49 (5.9s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 3,467 ms 2,395 ms 2,493 ms 226,803 ms 4,618 ms

首文本(SSE delta) - 2,475 ms 2,494 ms 226,806 ms 4,620 ms

总耗时 4,254 ms 2,500 ms 2,696 ms 227,300 ms 4,826 ms

输入 3 3 3 3 3

输出 5 5 5 5 5

缓存创建

32,323

5m: 0 / 1h: 32,323

31,989

5m: 0 / 1h: 31,989

32,318

5m: 0 / 1h: 32,318

32,318

5m: 0 / 1h: 32,318

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 14:43:27 ~ 14:43:37 (7.1s) 14:44:30 ~ 14:44:35 (4.5s) 14:45:38 ~ 14:45:43 (3.6s) 14:48:42 ~ 14:52:32 (225.8s) 14:43:51 ~ 14:43:56 (4.5s)

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。请只输出以下 3 行内容（label 小写，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) 3,140 ms 2,859 ms 2,235 ms 223,754 ms 3,121 ms

首文本(SSE delta) - 3,558 ms 2,236 ms 223,757 ms 3,123 ms

总耗时 4,520 ms 3,702 ms 2,638 ms 224,638 ms 3,391 ms

输入 3 3 3 3 3

输出 22 22 22 22 22

缓存创建

178

5m: 0 / 1h: 178

178

5m: 0 / 1h: 178

5m: 0 / 1h: 0

178

5m: 0 / 1h: 178

178

5m: 0 / 1h: 178

缓存读取 32,323 31,989 32,496 32,318 32,318

响应

vendor: anthropic
brand: claude
model: claude-sonnet-4-6

vendor: anthropic
brand: claude
model: claude-sonnet-4-5

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 14:45:08 ~ 14:45:18 (7.2s) 14:45:55 ~ 14:46:00 (3.5s) 14:48:03 ~ 14:48:08 (3.4s) 14:55:14 ~ 14:59:05 (228.9s) 14:47:01 ~ 14:47:06 (4.7s)

提示词

查看本步 Prompt

Knowledge cutoff date? Reply with YYYY-MM only.

HTTP 头(TTFB) 3,721 ms 2,387 ms 2,201 ms 227,325 ms 3,357 ms

首文本(SSE delta) - 2,559 ms 2,360 ms 227,338 ms 3,359 ms

总耗时 4,343 ms 2,669 ms 2,389 ms 227,858 ms 3,645 ms

输入 3 3 3 3 3

输出 8 8 8 8 8

缓存创建

5m: 0 / 1h: 38

5m: 0 / 1h: 57

5m: 0 / 1h: 0

5m: 0 / 1h: 55

缓存读取 32,501 32,167 32,551 32,496 32,496

响应

2025-08

2025-04

2025-08

2025-08

2025-08

步骤 4: identity_free text

时间 14:48:36 ~ 14:48:46 (7.8s) 14:48:27 ~ 14:48:32 (4.5s) 14:51:52 ~ 14:52:02 (5.6s) 15:01:15 ~ 15:02:55 (100.0s) 14:48:18 ~ 14:48:28 (5.2s)

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) 4,589 ms 3,373 ms 3,176 ms 97,810 ms 3,846 ms

首文本(SSE delta) - 3,637 ms 3,177 ms 97,811 ms 3,848 ms

总耗时 5,410 ms 3,793 ms 4,658 ms 98,796 ms 4,261 ms

输入 3 3 3 3 3

输出 59 24 44 50 19

缓存创建

32,574

5m: 0 / 1h: 32,574

5m: 0 / 1h: 35

5m: 0 / 1h: 39

5m: 0 / 1h: 43

5m: 0 / 1h: 35

缓存读取 0 32,224 32,551 32,551 32,551

输出速度 71.9 tok/s - 29.7 tok/s 50.8 tok/s -

响应

我是 Claude，由 Anthropic 开发的 AI 助手。我基于大型语言模型构建，旨在提供安全、有用且诚实的对话与任务协助。

我是 Claude，由 Anthropic 开发的 AI 助手。

我是 Claude，由 Anthropic 打造的 AI 助手。当前运行的模型版本为 claude-sonnet-4-6。

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。

我是Claude，由Anthropic开发的AI助手。

步骤 5: knowledge_recall text

时间 14:51:38 ~ 14:51:48 (7.0s) 14:50:36 ~ 14:50:46 (4.9s) 14:53:27 ~ 14:53:32 (4.5s) 15:04:17 ~ 15:05:17 (58.5s) 14:50:53 ~ 14:51:03 (5.5s)

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) 2,933 ms 1,824 ms 1,525 ms 54,583 ms 2,385 ms

首文本(SSE delta) - 3,939 ms 1,526 ms 54,584 ms 2,387 ms

总耗时 4,877 ms 4,108 ms 3,489 ms 57,557 ms 4,456 ms

输入 3 3 3 3 3

输出 205 198 205 204 205

缓存创建

8,017

5m: 0 / 1h: 8,017

7,672

5m: 0 / 1h: 7,672

2,305

5m: 0 / 1h: 2,305

8,033

5m: 0 / 1h: 8,033

7,994

5m: 0 / 1h: 7,994

缓存读取 0 0 5,718 0 0

输出速度 105.5 tok/s - 104.4 tok/s 68.6 tok/s 99.1 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

{"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 14:55:30 ~ 14:55:55 (21.6s) 14:54:30 ~ 14:54:35 (1.1s) 14:54:36 ~ 14:55:16 (38.6s) 15:08:23 ~ 15:08:43 (18.1s) 14:54:45 ~ 14:55:10 (20.8s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=231952805808099088792405976154953604840546894930954250263302023258694223

HTTP 头(TTFB) 1,567 ms - 2,207 ms 5,386 ms 1,601 ms

首文本(SSE delta) - - 2,209 ms 5,466 ms 1,603 ms

总耗时 20,440 ms - 37,641 ms 17,084 ms 19,807 ms

输入 164 - 164 164 164

输出 2,025 - 1,562 1,029 1,821

缓存创建

5m: 0 / 1h: 0

缓存读取 0 - 0 0 0

输出速度 107.3 tok/s - 44.1 tok/s 88.6 tok/s 100.0 tok/s

响应

## Counting digits in s

**s = `231952805808099088792405976154953604840546894930954250263302023258694223`**

### Step 1 — Total Length

Splitting into groups of 10:
```
2319528058 | 0809908879 | 24059

失败

评测失败

## Counting digits in s

**s** = `231952805808099088792405976154953604840546894930954250263302023258694223`

### Step-by-step indexing (1-based):

| Pos | Digit | | Pos | Digit | | Pos | Digit |
|----

```
s = "231952805808099088792405976154953604840546894930954250263302023258694223"
```

**Counting length:** stepping through all 72 characters.

**Digit-by-digit scan:**

| digit | positions (1-index

I'll count the digits in `s = 231952805808099088792405976154953604840546894930954250263302023258694223` by splitting into groups of 10:

| Group | Digits | 0s | 7s | 9s |
|---|---|---|---|---|
| `2319

总计

总输入(含缓存)

138,133

136,326

仅前 6 步

138,157

138,171

138,124

总输出

2,324

257

仅前 6 步

1,846

1,318

2,080

总缓存创建

73,130

5m: 0 / 1h: 73,130

39,931

5m: 0 / 1h: 39,931

仅前 6 步

34,662

5m: 0 / 1h: 34,662

40,627

5m: 0 / 1h: 40,627

40,580

5m: 0 / 1h: 40,580

总缓存读取

64,824

96,380

仅前 6 步

103,316

97,365

输出速度

105.8 tok/s

仅前 6 步

46.6 tok/s

82.4 tok/s

99.9 tok/s

总执行时间

56.5s

22.0s

仅前 6 步

59.6s

14m 19.5s

46.7s

总墙钟时间

14m 19.0s

含步骤间等待

13m 1.1s

含步骤间等待

仅前 6 步

13m 36.0s

含步骤间等待

27m 10.3s

含步骤间等待

13m 31.0s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.3291

Anthropic: Claude Sonnet 4.6

$0.1826

Anthropic: Claude Sonnet 4.6 (部分)

$0.1892

Anthropic: Claude Sonnet 4.6

$0.2019

Anthropic: Claude Sonnet 4.6

$0.2131

Anthropic: Claude Sonnet 4.6

通道指纹

来源步骤 digit_count knowledge_recall digit_count digit_count digit_count

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

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

CDN - - - - -

ID 格式 msg_01XKNUy7tRYCZ8321oE63L8F msg_015Q5rCjU6E4jzMGDHb8cwYQ msg_01CTMFgkLW88h3StKefoYBEY msg_0147MJk37xqNmJNnD4a4CeFj msg_01LsFknGFvGpVC8BkJAy3oeG

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	server: nginx/1.29.8 x-oneapi-request-id: 20260630065532610438358268d9d6vM1hKcub	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-idbody:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	—
O-Max-L2	cf-ray: a13b3ce38fd9f36b-LAX server: Photon-Edge set-cookie: _cfuvid=VOKtHnYpidsCZWZiOawc1dV_YHrrYZW_pwHpE3.V2jQ-1782802238.0100436-1.0.1.1-gq2L1mv_A8G_PoCd9ddvtoEFqM2LXYJK5LH_lh6Wjss; 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	body:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	1782802237898-8zjg2dfko04dl req_011CcYzy9WzHq6jdVapwHYbW a13b3ce38fd9f36b-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: 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_011CcZ1GqHdDfRXGnNUFCtMH
M-Max	server: nginx x-new-api-version: 0b8be5cf x-oneapi-request-id: 20260630150824163837816injDhLiI	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	—
O-Max	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_011CcZ1HSh7DUQGrepns6hQX

协议指纹评分（机器自动）

总分 / 100 88 100 95 99

基线状态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 5.5 /10 ×20 — 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 10.0 /10 ×15

模型匹配 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 10.0 /10 ×14 10.0 /10 ×14

sliding 5m 7.7 /10 ×13 — 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 10.0 /10 ×8

系统提示纯净 10.0 /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

知识截止 10.0 /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

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

inference_geo 10.0 /10 ×5 — 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5

延迟基线 9.5 /10 ×5 — 10.0 /10 ×5 0.0 /10 ×5 9.9 /10 ×5

Req-ID 透传 0.0 /10 ×4 — 10.0 /10 ×4 0.0 /10 ×4 10.0 /10 ×4

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

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

流式投递未计入 — 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5

综合结论

与基线相似度 88 v3.25.0 执行失败，不评分基准（参考） 95 v3.25.0 99 v3.25.0

雷达图例

共 19 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 自由身份 identity_free_clean
5 停止原因 stop_reason_present
6 系统提示纯净 system_prompt_clean
7 Service Tier service_tier_present
8 缓存命中比 cache_hit_ratio_match
9 缓存 TTL 一致 cache_ttl_consistency
10 推理地区 inference_geo_present
11 缓存连续 cache_continuity_intra
12 知识召回 knowledge_recall_match
13 延迟基线 latency_baseline_match
14 原生 msg-ID anthropic_msg_id_format
15 Sliding 缓存 cache_sliding_correctness
16 结构化身份 identity_structured_match
17 信封自报一致 envelope_self_report_match
18 档位思考量 tier_thinking_volume_match
19 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 7307 ms 3354 ms 8919 ms 142206 ms 6731 ms

输出速度 105.8 tok/s - 46.6 tok/s 82.4 tok/s 99.9 tok/s

完成步骤 6/6 成功 5/6 成功 · 1 失败 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 ⓘ

$0.3291

$0.1826 (部分)

$0.1892

$0.2019

$0.2131

缓存读取占比 ⓘ 47% 71% (部分) 75% 70% 70%

揭晓

通道

服务商 ByteCatCode ↗

通道O-Max

服务商 SSSAiCode ↗

通道O-Max-L2

官方基准

服务商 SudoCode ↗

通道M-Max

服务商 SAIAi ↗

通道O-Max

返回列表