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

2026/06/22 运行日期

claude-opus-4-8 官方基准

1 通道分数 < 70 另含 1 任务失败

已揭晓揭晓进度

步骤 / 指标

FoxCode

R-super

SSSAiCode

O-Max

SAIAi

O-Max

官方基准

Aiide

O-Max

基本信息

服务名 Claude Code Claude Code Claude Code Claude Code Claude Code

模型 claude-opus-4-8 claude-opus-4-8 claude-opus-4-8 claude-opus-4-8 claude-opus-4-8

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 18:07:55 ~ 18:08:00 (3.1s) 18:07:59 ~ 18:08:14 (10.1s) 18:08:01 ~ 18:08:06 (3.6s) 18:08:03 ~ 18:08:08 (3.4s) 18:07:57 ~ 18:08:07 (5.0s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 2,079 ms 7,925 ms 2,099 ms 2,138 ms 3,461 ms

首文本(SSE delta) 2,158 ms 8,001 ms 2,100 ms 2,140 ms 3,462 ms

总耗时 2,322 ms 8,073 ms 2,277 ms 2,264 ms 3,646 ms

输入 30 1,428 1,454 1,454 1,454

输出 2 4 4 4 4

缓存创建

23,529

5m: 23,529 / 1h: 0

29,761

5m: 0 / 1h: 29,761

5m: 0 / 1h: 0

30,380

5m: 0 / 1h: 30,380

缓存读取 0 0 30,222 30,222 0

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 18:10:29 ~ 18:10:35 (4.2s) 18:11:01 ~ 18:11:06 (2.9s) 18:09:30 ~ 18:09:35 (3.6s) 18:12:04 ~ 18:12:09 (3.6s) 18:10:15 ~ 18:10:20 (4.7s)

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) 1,892 ms 1,627 ms 1,875 ms 2,146 ms 3,549 ms

首文本(SSE delta) 1,894 ms 1,981 ms 1,877 ms 2,148 ms 3,550 ms

总耗时 3,485 ms 2,066 ms 2,271 ms 2,518 ms 3,554 ms

输入 76 2 2 2 2

输出 78 30 28 28 28

缓存创建

160

5m: 160 / 1h: 0

1,620

5m: 0 / 1h: 1,620

1,646

5m: 0 / 1h: 1,646

5m: 0 / 1h: 0

1,646

5m: 0 / 1h: 1,646

缓存读取 23,529 29,761 30,222 31,868 30,380

输出速度 49.0 tok/s - - - -

响应

vendor: anthropic
brand: claude
model: claude-opus-4-8

vendor: anthropic
brand: claude
model: claude-opus-4-20250514

vendor: anthropic
brand: claude
model: claude-opus-4-8

vendor: anthropic
brand: claude
model: claude-opus-4-8

vendor: anthropic
brand: claude
model: claude-opus-4-8

步骤 3: cutoff text

时间 18:11:50 ~ 18:11:55 (4.6s) 18:13:09 ~ 18:13:14 (3.4s) 18:10:51 ~ 18:10:56 (3.1s) 18:13:32 ~ 18:13:37 (4.1s) 18:11:36 ~ 18:11:41 (4.5s)

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) 2,912 ms 2,280 ms 1,825 ms 2,934 ms 3,224 ms

首文本(SSE delta) 2,913 ms 2,427 ms 1,826 ms 2,935 ms 3,225 ms

总耗时 3,339 ms 2,525 ms 2,048 ms 3,193 ms 3,439 ms

输入 76 2 2 2 2

输出 51 6 6 6 6

缓存创建

422

5m: 422 / 1h: 0

5m: 0 / 1h: 56

5m: 0 / 1h: 64

5m: 0 / 1h: 0

5m: 0 / 1h: 62

缓存读取 23,689 31,381 31,868 31,932 32,026

输出速度 119.7 tok/s - - - -

响应

2026-01

2025-01

2026-01

2026-01

2026-01

步骤 4: identity_free text

时间 18:15:33 ~ 18:15:38 (4.4s) 18:14:47 ~ 18:15:27 (35.4s) 18:13:18 ~ 18:13:23 (3.9s) 18:16:35 ~ 18:16:40 (3.6s) 18:13:20 ~ 18:13:30 (5.4s)

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) 2,522 ms 1,818 ms 1,821 ms 1,883 ms 2,945 ms

首文本(SSE delta) 2,523 ms 2,786 ms 1,823 ms 1,885 ms 2,946 ms

总耗时 3,206 ms 2,854 ms 2,893 ms 2,496 ms 4,080 ms

输入 79 2 2 2 2

输出 53 62 63 51 61

缓存创建

24,507

5m: 24,507 / 1h: 0

5m: 0 / 1h: 35

5m: 0 / 1h: 31

5m: 0 / 1h: 37

5m: 0 / 1h: 35

缓存读取 0 31,437 31,932 31,932 32,088

输出速度 77.6 tok/s - 58.9 tok/s 83.5 tok/s 53.8 tok/s

响应

我是 Claude，由 Anthropic 打造的 AI 助手，当前运行的模型为 claude-opus-4-8。

我是 Claude，由 Anthropic 公司打造的 AI 助手。具体为 Claude Opus 4 模型，可协助编程、写作、分析等多种任务。

我是 Claude，由 Anthropic 开发的人工智能助手，具体为 Opus 4.8 模型。我可以协助完成编程、写作、分析等多种任务。

我是 Claude，由 Anthropic 开发，属于 Claude 系列 AI 模型，具体为 Opus 4.8 版本。

我是 Claude，由 Anthropic 打造的人工智能助手。当前模型为 Claude Opus 4.8，可协助编程、写作、分析等多种任务。

步骤 5: knowledge_recall text

时间 18:18:16 ~ 18:18:51 (30.4s) 18:19:25 ~ 18:19:35 (5.8s) 18:16:40 ~ 18:16:50 (9.8s) 18:19:18 ~ 18:19:28 (8.3s) 18:17:24 ~ 18:17:39 (11.7s)

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) 1,951 ms 2,082 ms 3,829 ms 1,809 ms 4,966 ms

首文本(SSE delta) 1,952 ms 4,591 ms 6,629 ms 5,183 ms 7,160 ms

总耗时 29,437 ms 4,766 ms 8,873 ms 7,300 ms 10,513 ms

输入 32 2 2 2 2

输出 2,398 317 655 703 630

缓存创建

4,759

5m: 4,759 / 1h: 0

4,288

5m: 0 / 1h: 4,288

3,008

5m: 0 / 1h: 3,008

3,002

5m: 0 / 1h: 3,002

4,780

5m: 0 / 1h: 4,780

缓存读取 0 0 1,772 1,772 0

输出速度 87.2 tok/s - burst burst burst

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 18:19:55 ~ 18:20:10 (12.6s) 18:23:04 ~ 18:23:09 (1.3s) 18:20:04 ~ 18:20:09 (3.3s) 18:22:38 ~ 18:22:43 (3.8s) 18:21:02 ~ 18:21:22 (15.1s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=392153436375714250268965378234704388862594157322531528182325656663204101

HTTP 头(TTFB) 1,736 ms - 1,599 ms 2,511 ms 12,237 ms

首文本(SSE delta) 1,737 ms - 1,601 ms 2,512 ms 12,273 ms

总耗时 11,798 ms - 2,048 ms 2,871 ms 12,983 ms

输入 59 - 211 211 211

输出 1,133 - 19 19 19

缓存创建

149

5m: 149 / 1h: 0

5m: 0 / 1h: 0

缓存读取 0 - 0 0 0

输出速度 112.6 tok/s - - - -

响应

```
s = "392153436375714250268965378234704388862594157322531528182325656663204101"
```

Counting each digit position:

**Length:** 72

**0s** at indices 17, 31, 67, 70 → **4**

**7s** at indices 10, 1

失败

评测失败

{"n":72,"c":[3,5,3]}

{"n":72,"c":[5,5,3]}

{"n":72,"c":[5,7,6]}

总计

总输入(含缓存)

101,096

129,775

仅前 6 步

132,438

133,070

总输出

3,715

419

仅前 6 步

775

811

748

总缓存创建

53,526

5m: 53,526 / 1h: 0

35,760

5m: 0 / 1h: 35,760

仅前 6 步

4,749

5m: 0 / 1h: 4,749

3,039

5m: 0 / 1h: 3,039

36,903

5m: 0 / 1h: 36,903

总缓存读取

47,218

92,579

仅前 6 步

126,016

127,726

94,494

输出速度

92.3 tok/s

仅前 6 步

58.9 tok/s

83.5 tok/s

53.8 tok/s

总执行时间

59.4s

58.8s

仅前 6 步

27.3s

26.9s

46.4s

总墙钟时间

12m 15.1s

含步骤间等待

15m 10.1s

含步骤间等待

仅前 6 步

12m 8.0s

含步骤间等待

14m 39.9s

含步骤间等待

13m 25.0s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.3273

Anthropic: Claude Opus 4.8

$0.2874

Anthropic: Claude Opus 4.8 (部分)

$0.1204

Anthropic: Claude Opus 4.8

$0.1115

Anthropic: Claude Opus 4.8

$0.3050

Anthropic: Claude Opus 4.8

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

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

CDN cloudflare - - - cloudflare

ID 格式 msg_01OzUE7z1sLLv4JDzZAlYMA7 msg_01RmRvgHPUbeHcBimqpVERd1 msg_013a3P5hixQyz12snp64h1bo msg_01JxCMDg14pB8QfBme8GefFZ msg_01BKGFAUk5VSVVux49sCbqzU

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
R-super	cf-ray: a0fa847eb819e3af-NRT server: cloudflare x-new-api-version: v0.0.0 x-oneapi-request-id: 202606221019567526721768268d9d6gNtjzNWb	id_format: msg_prefix 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-idhdr:server:cloudflarehdr:cf-raybody:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsupstream:anthropic-direct-from-body	a0fa847eb819e3af-NRT
O-Max	cf-ray: a0fa83c40ff5d7af-LAX server: Photon-Edge set-cookie: _cfuvid=hNU6yzzB8RsEDeZJYdBG.Q13rShafpcyYS07lQK0Gak-1782123566.7298968-1.0.1.1-8X8EmjGRvmOx0586P55myl8R37iAyLevP.uQt2L6Vx0; 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	1782123566695-ej7k8k819j9gn req_011CcJ8RNTbP2dHNTa5dVViB a0fa83c40ff5d7af-LAX
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_011CcJ8UGV1Uf4Fw9bWBUPFR
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_011CcJ8fbZn3sN8pACX51MnR
O-Max	cf-ray: a0fa86299b420bc5-AMS server: cloudflare x-new-api-version: 0.12.11 x-oneapi-request-id: 202606221021051351313438268d9d6ECmmYTSI	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-idhdr:server:cloudflarehdr:cf-raybody:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	a0fa86299b420bc5-AMS

协议指纹评分（机器自动）

总分 / 100 72 100 100 95

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 5.3 /10 ×20 — 10.0 /10 ×20 10.0 /10 ×20 10.0 /10 ×20

缓存 TTL 一致 0.0 /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 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

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

流式投递 7.9 /10 ×5 — 10.0 /10 ×5 10.0 /10 ×5 2.0 /10 ×5

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

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

SDK 一致 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 72 v3.24.0 执行失败，不评分 100 v3.24.0 基准（参考） 95 v3.24.0

雷达图例

共 23 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 Thinking 存在 thinking_present
5 流式投递 buffer_dump_match
6 自由身份 identity_free_clean
7 停止原因 stop_reason_present
8 系统提示纯净 system_prompt_clean
9 Service Tier service_tier_present
10 缓存命中比 cache_hit_ratio_match
11 缓存 TTL 一致 cache_ttl_consistency
12 推理地区 inference_geo_present
13 缓存连续 cache_continuity_intra
14 知识召回 knowledge_recall_match
15 延迟基线 latency_baseline_match
16 原生 msg-ID anthropic_msg_id_format
17 Sliding 缓存 cache_sliding_correctness
18 结构化身份 identity_structured_match
19 自报身份一致 self_identity_consistency
20 信封自报一致 envelope_self_report_match
21 中文遵循 instruction_following_lang
22 档位思考量 tier_thinking_volume_match
23 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 8931 ms 4057 ms 3402 ms 3440 ms 6369 ms

输出速度 92.3 tok/s - 58.9 tok/s 83.5 tok/s 53.8 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 ⓘ

$0.3273

$0.2874 (部分)

$0.1204

$0.1115

$0.3050

缓存读取占比 ⓘ 47% 71% (部分) 95% 96% 71%

揭晓

通道

服务商 FoxCode ↗

通道R-super

服务商 SSSAiCode ↗

通道O-Max

服务商 SAIAi ↗

通道O-Max

官方基准

服务商 Aiide ↗

通道O-Max

返回列表