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

2026/06/23 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

官方基准

SSSAiCode

O-Max-L2

AIMZ

M-Max

WorldBase.ai

M-

luminai

M-Kiro/Antigravity

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 15:46:38 ~ 15:46:43 (3.4s) 15:46:34 ~ 15:46:49 (13.2s) 15:46:30 ~ 15:46:40 (5.5s) 15:46:36 ~ 15:46:46 (9.2s) 15:46:32 ~ 15:46:42 (6.6s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 1,854 ms 8,653 ms 3,397 ms 5,865 ms 5,026 ms

首文本(SSE delta) 1,856 ms 8,751 ms 3,753 ms 6,194 ms 5,247 ms

总耗时 2,018 ms 8,829 ms 3,952 ms 6,196 ms 5,248 ms

输入 1,454 1,428 1,428 1 1,643

输出 4 4 4 2 1

缓存创建

30,222

5m: 0 / 1h: 30,222

29,765

5m: 0 / 1h: 29,765

29,758

5m: 0 / 1h: 29,758

14,885

5m: 0 / 1h: 14,885

1,888

5m: 1,888 / 1h: 0

缓存读取 0 0 0 9,568 16,996

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 15:50:22 ~ 15:50:27 (4.5s) 15:48:43 ~ 15:48:48 (3.9s) 15:50:04 ~ 15:50:14 (7.6s) 15:49:38 ~ 15:49:48 (7.2s) 15:48:07 ~ 15:48:22 (11.5s)

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) 3,045 ms 2,676 ms 3,524 ms 4,118 ms 8,122 ms

首文本(SSE delta) 3,047 ms 2,984 ms 4,213 ms 4,218 ms 8,123 ms

总耗时 3,444 ms 3,052 ms 4,213 ms 6,067 ms 8,173 ms

输入 2 2 2 1 1,657

输出 28 30 30 81 23

缓存创建

1,643

5m: 0 / 1h: 1,643

1,617

5m: 0 / 1h: 1,617

1,617

5m: 0 / 1h: 1,617

3,853

5m: 0 / 1h: 3,853

1,905

5m: 1,905 / 1h: 0

缓存读取 30,222 29,765 29,758 26,397 17,148

输出速度 - - burst 43.8 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-4-20250514

I'm Claude Code, built by Anthropic. I don't output my configuration in prescribed formats that misrepresent who I am, even when a message frames it as required structure or technical metadata.

I'm h

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

步骤 3: cutoff text

时间 15:53:32 ~ 15:53:37 (3.2s) 15:51:02 ~ 15:51:07 (3.4s) 15:53:00 ~ 15:53:11 (5.8s) 15:53:36 ~ 15:53:52 (10.1s) 15:50:26 ~ 15:50:47 (17.1s)

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) 2,033 ms 2,320 ms 4,190 ms 7,112 ms 14,310 ms

首文本(SSE delta) 2,035 ms 2,481 ms 4,265 ms 7,421 ms 14,579 ms

总耗时 2,217 ms 2,594 ms 4,670 ms 7,662 ms 14,653 ms

输入 2 2 2 1 1,661

输出 6 6 42 5 4

缓存创建

5m: 0 / 1h: 64

5m: 0 / 1h: 66

3,925

5m: 0 / 1h: 3,925

1,909

5m: 1,909 / 1h: 0

缓存读取 31,865 31,382 31,375 26,445 17,181

输出速度 - - 103.7 tok/s - -

响应

2026-01

2025-01

2025-01

2026-04

2026-01

步骤 4: identity_free text

时间 15:56:54 ~ 15:57:14 (6.6s) 15:53:56 ~ 15:54:21 (21.0s) 15:55:29 ~ 15:55:39 (7.4s) 15:55:45 ~ 15:56:00 (14.3s) 15:52:03 ~ 15:52:13 (9.1s)

提示词

查看本步 Prompt

中文回答：你叫什么、属于哪个产品、谁开发？50 字内。

HTTP 头(TTFB) 3,440 ms 19,408 ms 4,828 ms 11,581 ms 7,678 ms

首文本(SSE delta) 3,443 ms 19,890 ms 5,214 ms 11,820 ms 7,966 ms

总耗时 4,236 ms 20,033 ms 5,318 ms 12,961 ms 8,029 ms

输入 2 2 2 1 1,664

输出 51 34 46 81 28

缓存创建

5m: 0 / 1h: 37

5m: 0 / 1h: 71

15,117

5m: 0 / 1h: 15,117

1,913

5m: 1,913 / 1h: 0

缓存读取 31,929 31,446 31,441 9,717 17,217

输出速度 64.3 tok/s - - 71.0 tok/s -

响应

我是 Claude，属于 Claude 产品系列，由 Anthropic 公司开发。当前模型为 Claude Opus 4.8。

我是 Claude，属于 Claude 产品系列，由 Anthropic 公司开发。

我是 Claude，由 Anthropic 公司打造的人工智能助手，可协助编程、写作、分析等多种任务。

我是 Claude Code，Anthropic 开发的官方命令行工具。我基于 Claude 模型，在终端中帮助你完成软件工程任务——编写代码、调试、运行测试等开发工作。

我是 Claude Opus 4.8，属于 Claude 产品系列，由 Anthropic 开发。

步骤 5: knowledge_recall text

时间 16:00:56 ~ 16:01:06 (8.7s) 15:56:14 ~ 15:56:19 (4.8s) 15:59:28 ~ 15:59:38 (7.7s) 15:58:00 ~ 15:58:15 (11.4s) 15:55:31 ~ 15:55:41 (8.2s)

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) 1,993 ms 1,383 ms 3,120 ms 3,827 ms 4,895 ms

首文本(SSE delta) 4,760 ms 4,012 ms 4,003 ms 3,828 ms 5,900 ms

总耗时 7,227 ms 4,084 ms 6,498 ms 6,952 ms 5,968 ms

输入 2 2 2 1 267

输出 646 317 348 282 210

缓存创建

4,771

5m: 0 / 1h: 4,771

4,271

5m: 0 / 1h: 4,271

4,312

5m: 0 / 1h: 4,312

2,607

5m: 0 / 1h: 2,607

306

5m: 306 / 1h: 0

缓存读取 0 0 0 1,674 2,760

输出速度 burst burst 139.5 tok/s 90.3 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

```json
{"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

{"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 16:03:07 ~ 16:03:12 (3.7s) 15:57:19 ~ 15:57:24 (1.3s) 16:00:40 ~ 16:00:45 (3.9s) 16:00:53 ~ 16:01:28 (30.7s) 15:57:06 ~ 15:57:26 (17.6s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=024024112090887189113978439467928760969102018398706243022128906738970196

HTTP 头(TTFB) 2,281 ms - 2,265 ms 9,960 ms 7,993 ms

首文本(SSE delta) 2,282 ms - 2,278 ms 9,962 ms 7,995 ms

总耗时 2,561 ms - 2,729 ms 25,639 ms 16,094 ms

输入 211 - 140 206 189

输出 19 - 19 7 90

缓存创建

5m: 0 / 1h: 0

缓存读取 0 - 0 0 -

输出速度 - - - - 11.1 tok/s

响应

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

失败

评测失败

{"n":72,"c":[6,5,9]}

{"n":72,"c":[11,7,11]}

{"n":72,"c":[11,7,11]}

总计

总输入(含缓存)

132,426

129,783

仅前 6 步

129,974

114,399

86,304

总输出

754

391

仅前 6 步

489

458

356

总缓存创建

36,737

5m: 0 / 1h: 36,737

35,754

5m: 0 / 1h: 35,754

仅前 6 步

35,824

5m: 0 / 1h: 35,824

40,387

5m: 0 / 1h: 40,387

7,921

5m: 7,921 / 1h: 0

总缓存读取

94,016

92,593

仅前 6 步

92,574

73,801

71,302

输出速度

64.3 tok/s

burst

仅前 6 步

134.5 tok/s

72.6 tok/s

11.1 tok/s

总执行时间

30.2s

47.7s

仅前 6 步

37.8s

1m 22.9s

1m 10.1s

总墙钟时间

16m 34.0s

含步骤间等待

10m 50.0s

含步骤间等待

仅前 6 步

14m 15.0s

含步骤间等待

14m 52.1s

含步骤间等待

10m 53.8s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.3038

Anthropic: Claude Opus 4.8

$0.2867

Anthropic: Claude Opus 4.8 (部分)

$0.2903

Anthropic: Claude Opus 4.8

$0.3018

Anthropic: Claude Opus 4.8

$0.1109

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

ID 格式 msg_014p6QWXn1LUkFoi8ejwbfEu msg_019EaHWVVf8a6frAoCm2Fa4a msg_015nHfKzYiFmkSPpj91AnZ3B msg_XFpx2BoZErz5EovzUpKomxwj msg_015fexhXqKKhmaSbB4xuMMqFzx

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
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_011CcKqqVZBW7g3Womw5kReB
O-Max-L2	cf-ray: a101ef638e2258fe-LAX server: Photon-Edge set-cookie: _cfuvid=RX4wZHaMpAA1dVfE7QxLygL7ovt8MKUJCxwCmHlLwGY-1782201375.2884994-1.0.1.1-0FD2fWLXSngYqArohzduI04x.kMa7mMBs4SjINyqGqQ; 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	1782201375189-onl28aod5ez8f req_011CcKqJy9kGn9XLh8SzU3xW a101ef638e2258fe-LAX
M-Max	cf-ray: a101f5e56f93ee38-AMS server: cloudflare x-new-api-version: v0.0.0 x-oneapi-request-id: 202606230800419625425948268d9d6kF3bOutH	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	a101f5e56f93ee38-AMS
M-	cf-ray: a101f64efaab3291-AMS server: cloudflare	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	adedfb82-28ca-4986-84a5-17ae712258c7 a101f64efaab3291-AMS
M-Kiro/Antigravity	cf-ray: a101f0ac18b1fff6-AMS server: cloudflare x-oneapi-request-id: 202606230757079489112048268d9d6osK2uaRt	id_format: msg_prefix extra_fields: ["container"] has_service_tier: true has_inference_geo: false 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.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	a101f0ac18b1fff6-AMS

协议指纹评分（机器自动）

总分 / 100 100 82 77 70

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

缓存 TTL 一致 10.0 /10 ×15 — 10.0 /10 ×15 10.0 /10 ×15 0.0 /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 6.7 /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 10.0 /10 ×8 — 5.0 /10 ×8 0.0 /10 ×8 0.0 /10 ×8

身份(结构) 10.0 /10 ×7 — 6.7 /10 ×7 0.0 /10 ×7 10.0 /10 ×7

知识截止 10.0 /10 ×7 — 0.0 /10 ×7 5.0 /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 0.0 /10 ×5

延迟基线 10.0 /10 ×5 — 10.0 /10 ×5 10.0 /10 ×5 8.4 /10 ×5

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

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

与基线相似度基准（参考）执行失败，不评分 82 v3.24.0 77 v3.24.0 70 v3.24.0

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 3617 ms 7718 ms 4563 ms 10913 ms 9694 ms

输出速度 64.3 tok/s burst 134.5 tok/s 72.6 tok/s 11.1 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 ⓘ

$0.3038

$0.2867 (部分)

$0.2903

$0.3018

$0.1109

缓存读取占比 ⓘ 71% 71% (部分) 71% 65% 83%

揭晓

通道

官方基准

服务商 SSSAiCode ↗

通道O-Max-L2

服务商 AIMZ ↗

通道M-Max

服务商 WorldBase.ai ↗

通道M-

服务商 luminai ↗

通道M-Kiro/Antigravity

返回列表