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

2 通道分数 < 70 与基线相似度

已揭晓揭晓进度

步骤 / 指标

0-0

O-Max

AiTokensFlux

O-Pro

OmniaKey

O-Max

SudoCode

M-Max

官方基准

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 00:55:27 ~ 00:55:37 (5.4s) 00:55:29 ~ 00:55:34 (4.3s) 00:55:25 ~ 00:55:35 (7.5s) 00:55:31 ~ 00:55:36 (3.6s) 00:55:33 ~ 00:55:38 (3.9s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 3,251 ms 2,795 ms 4,726 ms 2,090 ms 2,589 ms

首文本(SSE delta) 3,255 ms 2,797 ms 4,728 ms 2,098 ms 2,591 ms

总耗时 3,381 ms 2,931 ms 4,895 ms 2,273 ms 2,766 ms

输入 2 1,454 1,454 2 1,454

输出 4 4 4 4 4

缓存创建

32,303

5m: 32,303 / 1h: 0

30,380

5m: 0 / 1h: 30,380

30,222

5m: 0 / 1h: 30,222

31,672

5m: 0 / 1h: 31,672

186

5m: 0 / 1h: 186

缓存读取 0 0 0 0 30,036

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 00:58:41 ~ 00:58:46 (4.5s) 00:58:57 ~ 00:59:02 (3.2s) 00:59:31 ~ 00:59:41 (5.2s) 00:56:48 ~ 00:56:53 (3.6s) 00:57:48 ~ 00:57:53 (3.2s)

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。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) 2,764 ms 2,139 ms 3,682 ms 2,144 ms 1,795 ms

首文本(SSE delta) 2,765 ms 2,142 ms 3,684 ms 2,145 ms 1,797 ms

总耗时 3,117 ms 2,448 ms 4,168 ms 2,642 ms 2,181 ms

输入 2 2 2 2 2

输出 28 28 28 28 28

缓存创建

32,486

5m: 32,486 / 1h: 0

1,634

5m: 0 / 1h: 1,634

1,634

5m: 0 / 1h: 1,634

31,854

5m: 0 / 1h: 31,854

1,634

5m: 0 / 1h: 1,634

缓存读取 0 30,380 30,222 0 30,222

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 01:00:27 ~ 01:00:37 (8.1s) 01:00:55 ~ 01:01:05 (5.5s) 01:03:37 ~ 01:03:47 (5.7s) 00:59:19 ~ 00:59:24 (3.3s) 01:00:05 ~ 01:00:10 (4.0s)

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) 6,325 ms 4,354 ms 3,710 ms 1,997 ms 2,101 ms

首文本(SSE delta) 6,326 ms 4,372 ms 3,711 ms 2,009 ms 2,103 ms

总耗时 6,523 ms 4,591 ms 3,944 ms 2,368 ms 3,011 ms

输入 2 2 2 2 2

输出 6 6 6 6 6

缓存创建

32,549

5m: 32,549 / 1h: 0

5m: 0 / 1h: 54

5m: 0 / 1h: 62

31,916

5m: 0 / 1h: 31,916

5m: 0 / 1h: 54

缓存读取 0 32,014 31,856 0 31,856

响应

2026-01

2026-01

2026-01

2026-01

2026-01

步骤 4: identity_free text

时间 01:04:23 ~ 01:04:33 (8.4s) 01:03:40 ~ 01:03:45 (4.7s) 01:05:09 ~ 01:05:19 (6.8s) 01:01:41 ~ 01:01:46 (4.5s) 01:02:24 ~ 01:02:29 (3.8s)

提示词

查看本步 Prompt

请用中文介绍你自己是什么产品或助手，由谁开发，50 字以内。

HTTP 头(TTFB) 3,091 ms 2,307 ms 4,040 ms 1,898 ms 1,870 ms

首文本(SSE delta) 3,092 ms 2,308 ms 4,041 ms 1,899 ms 1,872 ms

总耗时 4,615 ms 3,292 ms 4,593 ms 3,078 ms 2,787 ms

输入 2 2 2 2 2

输出 80 67 61 67 61

缓存创建

32,589

5m: 32,589 / 1h: 0

5m: 0 / 1h: 31

5m: 0 / 1h: 39

31,955

5m: 0 / 1h: 31,955

5m: 0 / 1h: 35

缓存读取 0 32,068 31,918 0 31,910

输出速度 52.5 tok/s 68.1 tok/s 110.5 tok/s 56.8 tok/s 66.7 tok/s

响应

我是 Claude，由 Anthropic 开发的人工智能助手。我基于 Claude Opus 4.8 模型，可协助编程、写作、分析等任务，致力于提供有帮助、诚实且安全的回应。

我是 Claude，由 Anthropic 开发的人工智能助手，具体模型为 Claude Opus 4.8。我可以帮助你处理编程、写作、分析等各类任务。

我是 Claude，由 Anthropic 开发的 AI 助手，基于 Claude Opus 4.8 模型，能协助编程、写作、研究等多种任务。

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，致力于提供有帮助、诚实且安全的支持。

我是 Claude，由 Anthropic 打造的 AI 助手。当前模型为 Claude Opus 4.8，可用于编程、写作、研究等多种任务。

步骤 5: knowledge_recall text

时间 01:05:40 ~ 01:05:50 (9.2s) 01:06:51 ~ 01:07:01 (9.1s) 01:08:02 ~ 01:08:17 (10.4s) 01:03:51 ~ 01:04:01 (9.8s) 01:06:03 ~ 01:06:13 (8.3s)

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) 2,392 ms 1,888 ms 3,573 ms 3,348 ms 1,615 ms

首文本(SSE delta) 4,020 ms 4,370 ms 6,858 ms 6,674 ms 4,968 ms

总耗时 7,475 ms 8,184 ms 9,411 ms 8,883 ms 7,343 ms

输入 2 2 2 2 2

输出 660 702 684 683 688

缓存创建

5,376

5m: 5,376 / 1h: 0

4,762

5m: 0 / 1h: 4,762

4,772

5m: 0 / 1h: 4,772

4,776

5m: 0 / 1h: 4,776

4,760

5m: 0 / 1h: 4,760

缓存读取 0 0 0 0 0

输出速度 burst 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

{"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 01:07:47 ~ 01:07:52 (4.3s) 01:10:28 ~ 01:10:33 (3.3s) 01:11:39 ~ 01:11:49 (6.1s) 01:06:59 ~ 01:07:04 (3.0s) 01:09:31 ~ 01:09:36 (2.7s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=795092021292185971377237716655639051165792775761755182582303221965150256

HTTP 头(TTFB) 2,696 ms 2,069 ms 2,625 ms 1,105 ms 1,397 ms

首文本(SSE delta) 2,697 ms 2,070 ms 2,626 ms 1,106 ms 1,398 ms

总耗时 2,979 ms 2,477 ms 3,967 ms 1,857 ms 1,789 ms

输入 793 211 211 211 211

输出 19 19 19 19 19

缓存创建

5m: 0 / 1h: 0

缓存读取 0 0 0 0 0

响应

{"n":72,"c":[2,8,7]}

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

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

{"n":70,"c":[3,9,7]}

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

总计

总输入(含缓存) 136,106 132,996 132,398 132,394 132,366

总输出 797 826 802 807 806

总缓存创建

135,303

5m: 135,303 / 1h: 0

36,861

5m: 0 / 1h: 36,861

36,729

5m: 0 / 1h: 36,729

132,173

5m: 0 / 1h: 132,173

6,669

5m: 0 / 1h: 6,669

总缓存读取 0 94,462 93,996 0 124,024

输出速度 52.5 tok/s 68.1 tok/s 110.5 tok/s 56.8 tok/s 66.7 tok/s

总执行时间 39.9s 30.0s 41.7s 27.9s 25.9s

总墙钟时间

12m 24.9s

含步骤间等待

15m 4.0s

含步骤间等待

16m 24.0s

含步骤间等待

11m 32.9s

含步骤间等待

14m 2.9s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.5525

Anthropic: Claude Opus 4.8

$0.3066

Anthropic: Claude Opus 4.8

$0.3050

Anthropic: Claude Opus 4.8

$0.8474

Anthropic: Claude Opus 4.8

$0.1322

Anthropic: Claude Opus 4.8

通道指纹

来源步骤 digit_count digit_count digit_count digit_count digit_count

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

上游 Anthropic 直连 (100%) Anthropic 直连 (100%) Anthropic 直连 (100%) Anthropic 直连 (75%) Anthropic 直连 (100%)

CDN - - cloudflare - -

ID 格式 msg_01R74jFxmzqmamQVFymNnDNZ msg_017gBWf3HhGHfWr6YFfiBXYX msg_01ChJVAdBBa6ZXshcwHs6Wcg msg_vrtx_0167vJuBtHajKdkUo2UJqgWZ msg_019pBNVdDyqoRnN4uXqPfnUt

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	via: 1.1 Caddy server: uvicorn, nginx/1.24.0 (Ubuntu)	id_format: msg_prefix via_chain: ["1.1 Caddy"] extra_fields: ["stop_details"] has_service_tier: true has_inference_geo: true model_has_date_suffix: false usage_has_cache_fields: true	hdr:viabody:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	80df177f076828c701085ca69f34de24
O-Pro	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: a11597c7ce066622-AMS server: cloudflare x-oneapi-request-id: 202606251711425427976048268d9d6hu3luyGy	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	a11597c7ce066622-AMS
M-Max	server: nginx x-new-api-version: v0.0.0 x-oneapi-request-id: 2026062601070156411882aUTnHwJk	id_format: msg_prefix extra_fields: ["stop_details"] has_service_tier: false has_inference_geo: false model_has_date_suffix: false usage_has_cache_fields: true	hdr:x-new-api-versionhdr:x-oneapi-request-idbody:id:msg_prefixbody: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_011CcQM7wm77AZHm5VoCRcot

协议指纹评分（机器自动）

总分 / 100 65 95 98 68 100

基线状态已对照基线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 0.0 /10 ×20 10.0 /10 ×20 10.0 /10 ×20 0.0 /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 10.0 /10 ×15

模型匹配 10.0 /10 ×14 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 10.0 /10 ×14

sliding 5m 0.0 /10 ×13 10.0 /10 ×13 10.0 /10 ×13 0.0 /10 ×13 10.0 /10 ×13

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

系统提示纯净 5.0 /10 ×8 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 10.0 /10 ×7

知识截止 10.0 /10 ×7 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 10.0 /10 ×7

service_tier 10.0 /10 ×6 10.0 /10 ×6 10.0 /10 ×6 0.0 /10 ×6 10.0 /10 ×6

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

流式投递 0.0 /10 ×5 1.9 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5

Req-ID 透传 0.0 /10 ×4 0.0 /10 ×4 0.0 /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 10.0 /10 ×3

SDK 一致 10.0 /10 ×2 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 65 v3.24.1 95 v3.24.1 98 v3.24.1 68 v3.24.1 基准（参考）

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 4682 ms 3987 ms 5163 ms 3517 ms 3313 ms

输出速度 52.5 tok/s 68.1 tok/s 110.5 tok/s 56.8 tok/s 66.7 tok/s

完成步骤 6/6 成功 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 ⓘ

$0.5525

$0.3066

$0.3050

$0.8474

$0.1322

缓存读取占比 ⓘ 0% 71% 71% 0% 94%

揭晓

通道

服务商 0-0 ↗

通道O-Max

服务商 AiTokensFlux ↗

通道O-Pro

服务商 OmniaKey ↗

通道O-Max

服务商 SudoCode ↗

通道M-Max

官方基准

返回列表