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。

4 通道 · 6 步骤

2026/07/02 运行日期

claude-opus-4-8 官方基准

1 通道分数 < 70 另含 2 任务失败

已揭晓揭晓进度

⚠ 1 通道不可测 · 已从主对比表移除点击展开原因

TopRouterCN / O-Max
```
评测失败
```

步骤 / 指标

SudoCode

M-Max

官方基准

Aiide

O-Max

FoxCode

R-ultra

基本信息

服务名 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

客户端 claude-cli v2.1.195 claude-cli v2.1.195 claude-cli v2.1.195 claude-cli v2.1.195

偏差标红基于官方基准

步骤 1: ping text

时间 14:52:01 ~ 14:52:21 (16.9s) 14:52:07 ~ 14:52:12 (3.1s) 14:52:03 ~ 14:52:13 (6.7s) 14:51:59 ~ 14:52:04 (2.8s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 15,227 ms 1,970 ms 4,313 ms 1,759 ms

首文本(SSE delta) 15,237 ms 1,972 ms 4,315 ms 1,837 ms

总耗时 15,583 ms 2,072 ms 5,593 ms 2,005 ms

输入 2,326 2,326 2,326 14

输出 4 4 4 2

缓存创建

14,443

5m: 0 / 1h: 14,443

14,443

5m: 0 / 1h: 14,443

14,491

5m: 0 / 1h: 14,491

12,439

5m: 12,439 / 1h: 0

缓存读取 0 0 0 0

响应

pong

pong

pong

pong

步骤 2: identity text

时间 14:54:51 ~ 14:54:56 (4.0s) 14:53:45 ~ 14:53:50 (4.3s) 14:55:03 ~ 14:55:13 (5.0s) 14:54:43 ~ 14:56:59 (133.8s)

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) 2,651 ms 2,965 ms 2,884 ms 25,837 ms

首文本(SSE delta) 2,679 ms 2,967 ms 2,885 ms 81,984 ms

总耗时 2,858 ms 3,390 ms 3,171 ms -

输入 2 2 2 14

输出 28 28 28 74

缓存创建

16,949

5m: 0 / 1h: 16,949

2,506

5m: 0 / 1h: 2,506

2,506

5m: 0 / 1h: 2,506

12,634

5m: 12,634 / 1h: 0

缓存读取 0 14,443 14,491 0

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

失败

上游连接中断（未取得可评分响应）

步骤 3: cutoff text

时间 14:58:35 ~ 14:58:40 (3.9s) 14:57:34 ~ 14:57:39 (3.4s) 14:57:24 ~ 14:57:34 (7.1s) -

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) 2,540 ms 2,136 ms 5,789 ms -

首文本(SSE delta) 2,541 ms 2,137 ms 5,790 ms -

总耗时 2,792 ms 2,363 ms 6,020 ms -

输入 2 2 2 -

输出 6 6 6 -

缓存创建

5m: 0 / 1h: 62

5m: 0 / 1h: 64

5m: 0 / 1h: 62

缓存读取 16,949 16,949 16,997 -

响应

2026-01

2026-01

2026-01

未执行

步骤 4: identity_free text

时间 15:01:52 ~ 15:04:07 (135.0s) 15:01:26 ~ 15:01:32 (4.6s) 15:01:18 ~ 15:01:28 (8.9s) -

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) 133,238 ms 2,262 ms 4,462 ms -

首文本(SSE delta) 133,239 ms 2,264 ms 4,463 ms -

总耗时 134,031 ms 3,603 ms 7,846 ms -

输入 2 2 2 -

输出 65 71 51 -

缓存创建

2,607

5m: 0 / 1h: 2,607

5m: 0 / 1h: 39

5m: 0 / 1h: 31

缓存读取 14,443 17,013 17,059 -

输出速度 82.1 tok/s 53.0 tok/s 15.1 tok/s -

响应

我是 Claude，由 Anthropic 开发的 AI 助手。我能协助编程、写作、分析和解答各类问题，致力于提供有用、准确且安全的帮助。

我是 Claude，由 Anthropic 开发的 AI 助手。我可以帮你处理编程、写作、分析等各类任务，擅长软件工程与深入推理。很高兴为你服务！

我是 Claude，一个由 Anthropic 开发的人工智能助手。具体来说，我是 Claude Opus 4.8 模型。

未执行

步骤 5: knowledge_recall text

时间 15:06:42 ~ 15:08:47 (120.3s) 15:04:57 ~ 15:05:12 (10.3s) 15:05:12 ~ 15:05:27 (11.4s) -

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) 111,561 ms 2,364 ms 2,991 ms -

首文本(SSE delta) 117,004 ms 3,791 ms 5,474 ms -

总耗时 119,344 ms 9,402 ms 9,473 ms -

输入 2 2 2 -

输出 711 753 663 -

缓存创建

5,752

5m: 0 / 1h: 5,752

5,760

5m: 0 / 1h: 5,760

5,731

5m: 0 / 1h: 5,731

缓存读取 0 0 0 -

输出速度 burst 134.2 tok/s 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

未执行

步骤 6: digit_count text

时间 15:11:03 ~ 15:11:08 (3.3s) 15:06:36 ~ 15:06:41 (4.0s) 15:06:40 ~ 15:06:50 (5.5s) -

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=396715632428774357025386671389871380667473819002420073731100387032233352

HTTP 头(TTFB) 1,880 ms 2,608 ms 2,416 ms -

首文本(SSE delta) 1,881 ms 2,611 ms 3,022 ms -

总耗时 2,210 ms 3,049 ms 3,022 ms -

输入 0 211 211 -

输出 19 19 19 -

缓存创建

5m: 0 / 1h: 0

缓存读取 - 0 0 -

响应

{"n":72,"c":[9,11,4]}

{"n":72,"c":[6,11,3]}

{"n":72,"c":[7,12,4]}

未执行

总计

总输入(含缓存)

73,539

73,762

73,913

25,101

仅前 2 步

总输出

833

881

771

仅前 2 步

总缓存创建

39,813

5m: 0 / 1h: 39,813

22,812

5m: 0 / 1h: 22,812

22,821

5m: 0 / 1h: 22,821

25,073

5m: 25,073 / 1h: 0

仅前 2 步

总缓存读取

31,392

48,405

48,547

仅前 2 步

输出速度

82.1 tok/s

118.6 tok/s

15.1 tok/s

仅前 2 步

总执行时间

4m 43.5s

29.7s

44.6s

2m 16.6s

仅前 2 步

总墙钟时间

19m 7.4s

含步骤间等待

14m 33.9s

含步骤间等待

14m 47.0s

含步骤间等待

4m 59.5s

含步骤间等待

仅前 2 步

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

>=$0.2970

Anthropic: Claude Opus 4.8 (成本不完整)

$0.2015

Anthropic: Claude Opus 4.8

$0.1989

Anthropic: Claude Opus 4.8

$0.1000

Anthropic: Claude Opus 4.8 (部分)

通道指纹

来源步骤 digit_count digit_count digit_count identity

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

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

CDN - - cloudflare cloudflare

ID 格式自定义 msg_01BrKxigr5D8j5rbMSeCvT2v msg_01W3xyVbthwXuUKWujRctN4G msg_012XzJggodwclb5Mix0VgTm6

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-Max	server: nginx x-new-api-version: v1.0.0-rc.15 x-oneapi-request-id: 2026070215110561290765ym6PTGki	id_format: non_standard has_service_tier: false has_inference_geo: false model_has_date_suffix: false usage_has_cache_fields: false	hdr:x-new-api-versionhdr:x-oneapi-request-idbody:id:non_standard	—
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_011CccoohM7xhP8PkxwM4WBx
O-Max	cf-ray: a14bcf331811a00a-AMS server: cloudflare x-new-api-version: 0.12.11 x-oneapi-request-id: 202607020706432985085638268d9d6S4cZlMM0	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	a14bcf331811a00a-AMS
R-ultra	cf-ray: a14bbdab1f1efd4f-NRT server: cloudflare x-new-api-version: v0.0.0 x-oneapi-request-id: 202607020654507941553518268d9d6iBxby3Bu	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	a14bbdab1f1efd4f-NRT

协议指纹评分（机器自动）

总分 / 100 72 97 89

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 7.3 /10 ×20 7.3 /10 ×20 —

缓存 TTL 一致 10.0 /10 ×15 10.0 /10 ×15 10.0 /10 ×15 —

延迟基线 5.0 /10 ×15 10.0 /10 ×15 5.2 /10 ×15 —

模型匹配 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 —

sliding 5m 7.7 /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 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 —

身份(自由) 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 —

知识截止 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 —

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

流式投递 10.0 /10 ×5 10.0 /10 ×5 0.2 /10 ×5 —

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

Req-ID 透传 0.0 /10 ×2 10.0 /10 ×2 0.0 /10 ×2 —

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 72 v3.26.0 基准（参考） 89 v3.26.0 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 Thinking 体量 thinking_volume_match
14 缓存连续 cache_continuity_intra
15 知识召回 knowledge_recall_match
16 延迟基线 latency_baseline_match
17 原生 msg-ID anthropic_msg_id_format
18 Sliding 缓存 cache_sliding_correctness
19 结构化身份 identity_structured_match
20 信封自报一致 envelope_self_report_match
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 46136 ms 3980 ms 5854 ms 2005 ms

输出速度 82.1 tok/s 118.6 tok/s 15.1 tok/s -

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

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

>=$0.2970 (成本不完整)

$0.2015

$0.1989

$0.1000 (部分)

缓存读取占比 ⓘ 43% 66% 66% 0% (部分)

揭晓

通道

服务商 SudoCode ↗

通道M-Max

官方基准

服务商 Aiide ↗

通道O-Max

服务商 FoxCode ↗

通道R-ultra

返回列表