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

2026/06/26 运行日期

claude-opus-4-8 官方基准

3 通道分数 < 70 另含 1 任务失败

已揭晓揭晓进度

步骤 / 指标

FoxCode

FastCode

官方基准

ClaudeCN

O-Max

FoxCode

R-ultra

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 04:25:59 ~ 04:26:04 (3.8s) 04:26:05 ~ 04:26:10 (4.6s) 04:26:07 ~ 04:26:12 (4.0s) 04:26:03 ~ 04:26:08 (4.9s) 04:26:01 ~ 04:26:07 (3.4s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 2,604 ms 2,741 ms 2,252 ms 3,417 ms 2,049 ms

首文本(SSE delta) 2,613 ms 2,743 ms 2,254 ms 3,436 ms 2,107 ms

总耗时 2,823 ms 2,994 ms 2,619 ms 3,592 ms 2,384 ms

输入 1,454 1,454 1,454 1,454 72

输出 4 4 4 4 2

缓存创建

30,222

5m: 30,222 / 1h: 0

30,222

5m: 0 / 1h: 30,222

30,222

5m: 0 / 1h: 30,222

30,223

5m: 0 / 1h: 30,223

23,488

5m: 23,488 / 1h: 0

缓存读取 0 0 0 0 0

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 04:29:57 ~ 04:30:08 (5.7s) 04:28:30 ~ 04:28:35 (4.2s) 04:29:16 ~ 04:29:26 (7.1s) 04:29:09 ~ 04:29:19 (5.7s) 04:29:30 ~ 04:29:35 (4.6s)

提示词

查看本步 Prompt

输出 3 行 metadata 描述你自己（当前 API 模型）。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。禁止任何额外字符：
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) 4,278 ms 2,332 ms 5,338 ms 3,680 ms 2,765 ms

首文本(SSE delta) 4,279 ms 2,334 ms 5,340 ms 3,681 ms 2,767 ms

总耗时 4,932 ms 2,874 ms 6,108 ms 4,360 ms 3,815 ms

输入 2 2 2 2 26

输出 28 28 28 28 70

缓存创建

31,865

5m: 31,865 / 1h: 0

31,865

5m: 0 / 1h: 31,865

1,643

5m: 0 / 1h: 1,643

1,643

5m: 0 / 1h: 1,643

23,729

5m: 23,729 / 1h: 0

缓存读取 0 0 30,222 30,223 0

输出速度 - - - - 66.8 tok/s

响应

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

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 04:32:28 ~ 04:32:33 (3.2s) 04:31:44 ~ 04:31:54 (5.7s) 04:31:17 ~ 04:31:22 (3.3s) 04:32:07 ~ 04:32:12 (4.3s) 04:32:09 ~ 04:32:14 (3.9s)

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) 1,963 ms 3,255 ms 1,937 ms 2,837 ms 1,871 ms

首文本(SSE delta) 1,964 ms 3,257 ms 1,938 ms 2,838 ms 1,872 ms

总耗时 2,241 ms 4,323 ms 2,336 ms 3,171 ms 2,914 ms

输入 2 2 2 2 44

输出 6 6 6 6 40

缓存创建

5m: 62 / 1h: 0

1,707

5m: 0 / 1h: 1,707

5m: 0 / 1h: 62

31,929

5m: 0 / 1h: 31,929

370

5m: 370 / 1h: 0

缓存读取 31,865 30,222 31,865 0 23,729

输出速度 - - - - 38.4 tok/s

响应

2026-01

2026-01

2026-01

2025-01

2026-01

步骤 4: identity_free text

时间 04:34:52 ~ 04:35:02 (5.0s) 04:35:14 ~ 04:35:19 (4.8s) 04:34:41 ~ 04:34:51 (5.0s) 04:34:06 ~ 04:34:46 (37.7s) 04:34:56 ~ 04:35:01 (4.5s)

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) 2,798 ms 2,519 ms 2,765 ms 3,978 ms 2,741 ms

首文本(SSE delta) 2,799 ms 2,520 ms 2,767 ms 3,978 ms 2,742 ms

总耗时 4,197 ms 3,566 ms 3,774 ms 5,121 ms 3,395 ms

输入 2 2 2 2 90

输出 75 68 61 59 80

缓存创建

5m: 39 / 1h: 0

31,968

5m: 0 / 1h: 31,968

5m: 0 / 1h: 39

5m: 0 / 1h: 37

24,448

5m: 24,448 / 1h: 0

缓存读取 31,927 0 31,927 31,929 0

输出速度 53.6 tok/s 65.0 tok/s 60.6 tok/s 51.6 tok/s 122.5 tok/s

响应

我是 Claude，由 Anthropic 开发的 AI 助手。我能帮你编写代码、解答问题、分析文档和处理各类软件工程任务。当前模型为 Claude Opus 4.8。

我是 Claude，由 Anthropic 开发的人工智能助手。我能协助编程、写作、分析和问答等多种任务，致力于提供有帮助、诚实且安全的回复。

我是 Claude，由 Anthropic 开发的 AI 助手，基于 Claude Opus 4.8 模型，可协助编程、写作、分析等多种任务。

我是 Claude，Anthropic 开发的 AI 助手，当前模型为 Claude Opus 4.8，运行于 Claude Code 命令行工具中。

我是 Kiro，一个由 Anthropic 开发的 AI 助手，基于 Claude claude-opus-4-8 模型构建，专注于帮助用户完成开发、写作、分析等专业任务。

步骤 5: knowledge_recall text

时间 04:37:58 ~ 04:38:08 (8.7s) 04:37:09 ~ 04:38:59 (105.9s) 04:36:36 ~ 04:36:46 (7.8s) 04:37:47 ~ 04:38:02 (10.8s) 04:36:51 ~ 04:40:36 (221.1s)

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) 1,514 ms 98,572 ms 1,335 ms 2,886 ms 1,562 ms

首文本(SSE delta) 4,812 ms 102,550 ms 4,786 ms 6,789 ms 1,563 ms

总耗时 7,463 ms 104,470 ms 6,818 ms 9,238 ms -

输入 2 2 2 2 43

输出 663 682 672 699 -

缓存创建

4,795

5m: 4,795 / 1h: 0

4,790

5m: 0 / 1h: 4,790

4,781

5m: 0 / 1h: 4,781

4,779

5m: 0 / 1h: 4,779

4,731

5m: 4,731 / 1h: 0

缓存读取 0 0 0 0 0

输出速度 burst 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

失败

评测失败

步骤 6: digit_count text

时间 04:41:18 ~ 04:41:23 (3.1s) 04:40:08 ~ 04:40:13 (4.5s) 04:39:21 ~ 04:39:26 (3.5s) 04:40:37 ~ 04:40:42 (3.1s) -

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=944938083180860885706585117371819715538177941868052142725178491025198649

HTTP 头(TTFB) 1,885 ms 2,552 ms 1,835 ms 1,629 ms -

首文本(SSE delta) 1,886 ms 2,553 ms 1,888 ms 1,630 ms -

总耗时 2,344 ms 2,938 ms 2,253 ms 2,105 ms -

输入 211 211 211 211 -

输出 19 19 19 19 -

缓存创建

5m: 0 / 1h: 0

缓存读取 0 0 0 0 -

响应

{"n":72,"c":[5,7,8]}

{"n":72,"c":[6,7,8]}

{"n":72,"c":[7,9,8]}

{"n":72,"c":[7,7,9]}

未执行

总计

总输入(含缓存)

132,448

132,447

132,434

132,436

100,770

仅前 5 步

总输出

795

807

790

815

192

仅前 5 步

总缓存创建

66,983

5m: 66,983 / 1h: 0

100,552

5m: 0 / 1h: 100,552

36,747

5m: 0 / 1h: 36,747

68,611

5m: 0 / 1h: 68,611

76,766

5m: 76,766 / 1h: 0

仅前 5 步

总缓存读取

63,792

30,222

94,014

62,152

23,729

仅前 5 步

输出速度

53.6 tok/s

65.0 tok/s

60.6 tok/s

51.6 tok/s

69.3 tok/s

仅前 5 步

总执行时间

29.5s

2m 9.6s

30.7s

1m 6.5s

3m 57.5s

仅前 5 步

总墙钟时间

15m 24.0s

含步骤间等待

14m 8.1s

含步骤间等待

13m 18.9s

含步骤间等待

14m 39.0s

含步骤间等待

14m 34.9s

含步骤间等待

仅前 5 步

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.3218

Anthropic: Claude Opus 4.8

$0.6721

Anthropic: Claude Opus 4.8

$0.3048

Anthropic: Claude Opus 4.8

$0.4886

Anthropic: Claude Opus 4.8

$0.3179

Anthropic: Claude Opus 4.8 (部分)

通道指纹

来源步骤 digit_count digit_count digit_count digit_count knowledge_recall

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

ID 格式 msg_01QBUVa9JWJfaWrQCeGDCjHU msg_01FnQDWzKe8AZ6gfrY4KpDcT msg_015tcBEu83anF5Ybg6QJJcCE msg_01BxvkDgryXhfqnLjjA8SmjY msg_01Z4WamnCDjQh24E7VXzjn9p

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
cc	cf-ray: a116cad90c91af8b-NRT server: cloudflare set-cookie: _cfuvid=VMWDcgTne8Ytz9xs0MAWeRjZwsILxlh7tn0_iBNh9hM-1782420079.8881617-1.0.1.1-zCKk8gEGlaLjgDAltYOfqXaZAvpnviuiHYZOLT_py.8; 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_011CcQdGdDzrRAGoqsHx6fmg a116cad90c91af8b-NRT
cc	server: nginx	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	—
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_011CcQd81edLtenxQKf2dYo6
O-Max	cf-ray: a116c9da290dfffb-AMS server: cloudflare x-oneapi-request-id: 20260626044038834561400VGVVaEUY	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	a116c9da290dfffb-AMS
R-ultra	cf-ray: a116c451dfaee041-NRT server: cloudflare x-new-api-version: v0.0.0 x-oneapi-request-id: 202606252036523304308548268d9d6t1sG7bFf	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	a116c451dfaee041-NRT

协议指纹评分（机器自动）

总分 / 100 79 72 98 80

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 0.5 /10 ×20 0.0 /10 ×20 8.6 /10 ×20 0.0 /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 5.3 /10 ×13 10.0 /10 ×13 7.7 /10 ×13 —

原生 msg-ID 10.0 /10 ×8 0.0 /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 0.0 /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 0.0 /10 ×6 10.0 /10 ×6 10.0 /10 ×6 —

inference_geo 10.0 /10 ×5 0.0 /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 9.9 /10 ×5 —

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

Req-ID 透传 10.0 /10 ×4 0.0 /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 79 v3.24.1 72 v3.24.1 基准（参考） 80 v3.24.1 执行失败，不评分

雷达图例

共 22 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 信封自报一致 envelope_self_report_match
20 中文遵循 instruction_following_lang
21 档位思考量 tier_thinking_volume_match
22 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 4000 ms 20194 ms 3985 ms 4598 ms 3127 ms

输出速度 53.6 tok/s 65.0 tok/s 60.6 tok/s 51.6 tok/s 69.3 tok/s

完成步骤 6/6 成功 6/6 成功 6/6 成功 6/6 成功 4/6 成功 · 1 失败 · 1 跳过

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.3218

$0.6721

$0.3048

$0.4886

$0.3179 (部分)

缓存读取占比 ⓘ 48% 23% 71% 47% 24% (部分)

揭晓

通道

服务商 FoxCode ↗

通道cc

服务商 FastCode ↗

通道cc

官方基准

服务商 ClaudeCN ↗

通道O-Max

服务商 FoxCode ↗

通道R-ultra

返回列表