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

2026/06/30 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

⚠ 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

ClaudeZZ

R-kiro

官方基准

AnyAIGC

M-

Aiide

O-Max

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 18:00:40 ~ 18:00:50 (5.3s) 18:00:44 ~ 18:00:49 (4.2s) 18:00:38 ~ 18:00:48 (7.5s) 18:00:42 ~ 18:00:52 (5.0s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 3,116 ms 1,984 ms 2,761 ms 3,160 ms

首文本(SSE delta) 3,118 ms 1,985 ms 2,763 ms 3,163 ms

总耗时 3,587 ms 2,100 ms 2,767 ms 3,498 ms

输入 2 2,326 2,326 2,326

输出 4 4 4 4

缓存创建

16,029

5m: 0 / 1h: 16,029

1,711

5m: 0 / 1h: 1,711

14,491

5m: 0 / 1h: 14,491

186

5m: 0 / 1h: 186

缓存读取 0 12,732 0 14,305

响应

pong

pong

pong

pong

步骤 2: identity text

时间 18:03:33 ~ 18:03:38 (4.8s) 18:04:28 ~ 18:04:33 (3.6s) 18:02:20 ~ 18:02:30 (7.0s) 18:03:05 ~ 18:03:16 (5.2s)

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。严格按以下 3 行格式输出，全部小写英文，不要任何其他文字、解释、标点：
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) 3,608 ms 2,273 ms 4,511 ms 3,790 ms

首文本(SSE delta) 3,609 ms 2,275 ms 4,512 ms 3,791 ms

总耗时 4,031 ms 2,666 ms 4,673 ms 4,155 ms

输入 2 2 2 2

输出 59 28 28 28

缓存创建

203

5m: 0 / 1h: 203

2,515

5m: 0 / 1h: 2,515

2,515

5m: 0 / 1h: 2,515

2,515

5m: 0 / 1h: 2,515

缓存读取 16,029 14,443 14,491 14,491

输出速度 139.8 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

步骤 3: cutoff text

时间 18:05:41 ~ 18:05:51 (5.5s) 18:05:34 ~ 18:05:39 (3.6s) 18:04:33 ~ 18:04:38 (3.8s) 18:06:55 ~ 18:07:00 (4.4s)

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) 4,391 ms 2,337 ms 2,483 ms 2,849 ms

首文本(SSE delta) 4,392 ms 2,339 ms 2,484 ms 2,850 ms

总耗时 4,762 ms 2,588 ms 2,608 ms 3,273 ms

输入 2 2 2 2

输出 6 6 6 6

缓存创建

131

5m: 0 / 1h: 131

5m: 0 / 1h: 62

5m: 0 / 1h: 54

5m: 0 / 1h: 64

缓存读取 16,232 16,958 17,006 17,006

响应

2025-03

2026-01

2026-01

2026-01

步骤 4: identity_free text

时间 18:09:51 ~ 18:10:01 (5.7s) 18:07:56 ~ 18:08:06 (5.0s) 18:06:36 ~ 18:06:46 (9.3s) 18:08:32 ~ 18:08:42 (6.2s)

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) 3,576 ms 2,720 ms 6,321 ms 2,798 ms

首文本(SSE delta) 3,578 ms 2,722 ms 6,322 ms 2,843 ms

总耗时 4,881 ms 3,976 ms 8,070 ms 5,156 ms

输入 2 2 2 2

输出 69 61 47 88

缓存创建

5m: 0 / 1h: 46

5m: 0 / 1h: 31

5m: 0 / 1h: 37

5m: 0 / 1h: 39

缓存读取 16,363 17,020 17,060 17,070

输出速度 53.0 tok/s 48.6 tok/s 26.9 tok/s 38.0 tok/s

响应

我是 Claude，由 Anthropic 开发的 AI 助手。当前模型是 Claude Opus 4.8，可通过 Claude Code 在终端中协助你完成编程等任务。

我是 Claude，一个由 Anthropic 开发的人工智能助手。具体来说，我运行的模型是 Opus 4.8（claude-opus-4-8）。

我是 Claude，由 Anthropic 开发，属于 Claude 系列模型，当前为 Opus 4.8 版本。

我是 Claude，由 Anthropic 开发的 AI 助手，基于 Claude Agent SDK 构建，可协助软件工程、研究分析等任务。当前模型为 Opus 4.8，擅长编程与多步骤任务处理。

步骤 5: knowledge_recall text

时间 18:12:34 ~ 18:12:49 (10.0s) 18:11:34 ~ 18:11:44 (9.2s) 18:09:33 ~ 18:09:48 (13.0s) 18:11:13 ~ 18:11:28 (10.1s)

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) 3,441 ms 1,542 ms 4,077 ms 3,251 ms

首文本(SSE delta) 6,334 ms 2,959 ms 6,161 ms 4,811 ms

总耗时 9,126 ms 8,346 ms 11,788 ms 9,094 ms

输入 2 2 2 2

输出 669 734 752 703

缓存创建

12,048

5m: 0 / 1h: 12,048

5,749

5m: 0 / 1h: 5,749

5,734

5m: 0 / 1h: 5,734

5,786

5m: 0 / 1h: 5,786

缓存读取 0 0 0 0

输出速度 burst 136.3 tok/s 133.6 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

{"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 18:15:06 ~ 18:15:21 (12.7s) 18:13:15 ~ 18:13:20 (3.2s) 18:11:22 ~ 18:11:27 (4.1s) 18:13:55 ~ 18:14:00 (3.6s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=704317031950037680795818837933234879050871160251015997410413842865521111

HTTP 头(TTFB) 4,035 ms 1,707 ms 1,835 ms 2,055 ms

首文本(SSE delta) 11,431 ms 1,709 ms 1,836 ms 2,056 ms

总耗时 11,983 ms 2,062 ms 1,988 ms 2,387 ms

输入 2 211 211 211

输出 931 19 19 19

缓存创建

5,089

5m: 0 / 1h: 5,089

5m: 0 / 1h: 0

缓存读取 3,067 0 0 0

输出速度 burst - - -

响应

{"n":72,"c":[10,8,6]}

{"n":72,"c":[10,9,7]}

{"n":72,"c":[12,7,5]}

{"n":72,"c":[8,9,5]}

总计

总输入(含缓存) 85,249 73,766 73,933 74,007

总输出 1,738 852 856 848

总缓存创建

33,546

5m: 0 / 1h: 33,546

10,068

5m: 0 / 1h: 10,068

22,831

5m: 0 / 1h: 22,831

8,590

5m: 0 / 1h: 8,590

总缓存读取 51,691 61,153 48,557 62,872

输出速度 74.2 tok/s 119.7 tok/s 108.3 tok/s 38.0 tok/s

总执行时间 43.9s 28.8s 44.8s 34.6s

总墙钟时间

14m 41.1s

含步骤间等待

12m 36.0s

含步骤间等待

10m 49.4s

含步骤间等待

13m 18.2s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.2790

Anthropic: Claude Opus 4.8

$0.1275

Anthropic: Claude Opus 4.8

$0.2011

Anthropic: Claude Opus 4.8

$0.1190

Anthropic: Claude Opus 4.8

通道指纹

来源步骤 digit_count digit_count digit_count digit_count

平台 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%)

CDN - - - cloudflare

ID 格式 msg_018k9tEz6geUX3bHo3wr66tS msg_0153yev6TcHE6cPU8f5RMC4p msg_01B5tGzv2dsqnYKQGHKAhdc5 msg_016dkWhCNTBZ6bKLoXE1o1LN

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-kiro	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_011CcZGRRRvbVZqfJXQNw8aD
M-	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	—
O-Max	cf-ray: a13c66b75ad793c0-AMS server: cloudflare x-new-api-version: 0.12.11 x-oneapi-request-id: 202606301013572468570668268d9d6vaxaxxxM	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	a13c66b75ad793c0-AMS

协议指纹评分（机器自动）

总分 / 100 79 100 94 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 8.6 /10 ×20 10.0 /10 ×20 9.7 /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

延迟基线 9.4 /10 ×15 10.0 /10 ×15 8.8 /10 ×15 9.8 /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 10.0 /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

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

知识截止 0.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5

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

stop_reason 10.0 /10 ×3 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 0.0 /10 ×2

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

流式投递未计入未计入未计入未计入

综合结论

与基线相似度 79 v3.26.0 基准（参考） 94 v3.26.0 95 v3.26.0

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 Thinking 存在 thinking_present
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 Thinking 体量 thinking_volume_match
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 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 6395 ms 3623 ms 5316 ms 4594 ms

输出速度 74.2 tok/s 119.7 tok/s 108.3 tok/s 38.0 tok/s

完成步骤 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 ⓘ

$0.2790

$0.1275

$0.2011

$0.1190

缓存读取占比 ⓘ 61% 83% 66% 85%

揭晓

通道

服务商 ClaudeZZ ↗

通道R-kiro

官方基准

服务商 AnyAIGC ↗

通道M-

服务商 Aiide ↗

通道O-Max

返回列表