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

2026/06/18 运行日期

claude-opus-4-8 官方基准

2 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

DawAPI / cc
```
评测失败
```

步骤 / 指标

官方基准

SSSAiCode

O-Max-L2

Novaris

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 v2.1.160 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 15:58:34 ~ 15:58:39 (4.1s) 15:58:28 ~ 15:58:33 (4.2s) 15:58:30 ~ 15:58:40 (7.2s) 15:58:26 ~ 15:58:31 (3.9s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 2,794 ms 2,765 ms 6,117 ms 2,632 ms

首文本(SSE delta) 2,796 ms 2,838 ms 6,174 ms 2,675 ms

总耗时 2,967 ms 2,911 ms 6,379 ms 2,905 ms

输入 1,454 1,428 1,484 50

输出 4 4 4 2

缓存创建

30,222

5m: 0 / 1h: 30,222

29,765

5m: 0 / 1h: 29,765

27,567

5m: 27,567 / 1h: 0

23,510

5m: 23,510 / 1h: 0

缓存读取 0 0 0 0

响应

pong

pong

pong

pong

步骤 2: identity text

时间 16:00:05 ~ 16:00:10 (4.8s) 15:59:36 ~ 15:59:46 (5.2s) 15:59:45 ~ 15:59:55 (6.0s) 16:01:36 ~ 16:01:41 (4.5s)

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。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) 3,310 ms 3,595 ms 4,683 ms 2,385 ms

首文本(SSE delta) 3,312 ms 4,022 ms 4,734 ms 2,386 ms

总耗时 3,802 ms 4,154 ms 5,157 ms 3,650 ms

输入 2 2 2 28

输出 28 32 28 70

缓存创建

1,634

5m: 0 / 1h: 1,634

1,608

5m: 0 / 1h: 1,608

1,664

5m: 1,664 / 1h: 0

23,727

5m: 23,727 / 1h: 0

缓存读取 30,222 29,765 27,567 0

输出速度 - - - 55.4 tok/s

响应

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

vendor: anthropic
brand: claude
model: claude-opus-4-5-20251101

vendor: anthropic
brand: claude
model: claude-opus-4-1

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

步骤 3: cutoff text

时间 16:02:30 ~ 16:02:35 (4.0s) 16:01:06 ~ 16:01:16 (6.0s) 16:02:44 ~ 16:02:54 (6.1s) 16:03:14 ~ 16:03:19 (4.0s)

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) 2,754 ms 1,900 ms 5,002 ms 2,229 ms

首文本(SSE delta) 2,756 ms 2,065 ms 5,003 ms 2,230 ms

总耗时 2,986 ms 2,210 ms 5,318 ms 3,179 ms

输入 2 2 2 75

输出 6 6 6 37

缓存创建

5m: 0 / 1h: 62

5m: 0 / 1h: 68

5m: 54 / 1h: 0

345

5m: 345 / 1h: 0

缓存读取 31,856 31,373 29,231 23,727

输出速度 - - - 39.0 tok/s

响应

2026-01

2025-03

2025-01

2026-01

步骤 4: identity_free text

时间 16:03:59 ~ 16:04:04 (3.7s) 16:03:40 ~ 16:03:45 (3.1s) 16:05:34 ~ 16:05:39 (4.3s) 16:07:13 ~ 16:07:18 (4.8s)

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) 2,194 ms 1,605 ms 2,518 ms 2,012 ms

首文本(SSE delta) 2,196 ms 2,286 ms 2,519 ms 2,013 ms

总耗时 2,737 ms 2,348 ms 3,610 ms 3,904 ms

输入 2 2 2 29

输出 51 45 71 88

缓存创建

5m: 0 / 1h: 31

5m: 0 / 1h: 37

5m: 39 / 1h: 0

24,505

5m: 24,505 / 1h: 0

缓存读取 31,918 31,441 29,285 0

输出速度 94.3 tok/s - 65.1 tok/s 46.5 tok/s

响应

我是 Claude，一个由 Anthropic 开发的 AI 助手。具体来说，我是 Claude Opus 4.8 模型。

我是 Claude，由 Anthropic 开发，属于 Claude 系列 AI 助手中的 Opus 模型。

我是 Claude Code，由 Anthropic 开发的官方命令行（CLI）AI 助手，基于 Claude 模型，帮助你在终端中编写代码、执行任务和管理项目。

我是 Kiro，一款由 Anthropic 的 Claude 模型驱动的 AI 编程助手，帮助开发者完成软件工程任务。

步骤 5: knowledge_recall text

时间 16:07:57 ~ 16:08:07 (8.5s) 16:06:40 ~ 16:06:50 (7.8s) 16:07:43 ~ 16:07:53 (6.0s) 16:10:44 ~ 16:11:04 (18.7s)

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) 2,077 ms 4,491 ms 2,571 ms 1,641 ms

首文本(SSE delta) 4,920 ms 4,493 ms 2,635 ms 1,642 ms

总耗时 7,392 ms 7,126 ms 5,239 ms 17,868 ms

输入 2 2 2 76

输出 686 344 344 1,398

缓存创建

4,754

5m: 0 / 1h: 4,754

4,277

5m: 0 / 1h: 4,277

3,059

5m: 3,059 / 1h: 0

4,691

5m: 4,691 / 1h: 0

缓存读取 0 0 0 0

输出速度 burst 130.6 tok/s 132.1 tok/s 86.2 tok/s

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 16:11:23 ~ 16:11:28 (2.9s) 16:08:38 ~ 16:08:43 (2.8s) 16:11:52 ~ 16:12:02 (8.2s) 16:12:57 ~ 16:13:12 (12.2s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=216393454158079436554457084332490307995836758695030826967948942152630192

HTTP 头(TTFB) 1,631 ms - 1,585 ms 2,105 ms

首文本(SSE delta) 1,634 ms - 7,242 ms 2,106 ms

总耗时 1,970 ms - 7,443 ms 11,418 ms

输入 211 - 193 35

输出 19 - 737 1,112

缓存创建

5m: 0 / 1h: 0

173

5m: 173 / 1h: 0

缓存读取 0 - 0 0

输出速度 - - burst 119.4 tok/s

响应

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

失败

评测失败

{"n":72,"c":[7,5,10]}

{"n":72,"c":[7,5,10]}

总计

总输入(含缓存)

132,372

129,770

仅前 6 步

120,151

100,971

总输出

794

431

仅前 6 步

1,190

2,707

总缓存创建

36,703

5m: 0 / 1h: 36,703

35,755

5m: 0 / 1h: 35,755

仅前 6 步

32,383

5m: 32,383 / 1h: 0

76,951

5m: 76,951 / 1h: 0

总缓存读取

93,996

92,579

仅前 6 步

86,083

23,727

输出速度

94.3 tok/s

130.6 tok/s

仅前 6 步

112.3 tok/s

91.3 tok/s

总执行时间

28.1s

29.1s

仅前 6 步

37.8s

48.1s

总墙钟时间

12m 54.0s

含步骤间等待

10m 15.0s

含步骤间等待

仅前 6 步

13m 31.9s

含步骤间等待

14m 46.0s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.3046

Anthropic: Claude Opus 4.8

$0.2877

Anthropic: Claude Opus 4.8 (部分)

$0.2077

Anthropic: Claude Opus 4.8

$0.3816

Anthropic: Claude Opus 4.8

通道指纹

来源步骤 digit_count knowledge_recall digit_count digit_count

平台 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%)

CDN - - - cloudflare

ID 格式 msg_01UNVVBbHXw3hS8txzjdSGps msg_01HxzdRanX9nHMT5LB6Y7UdN msg_01BChXfdb33qf5P3sfoANXZu msg_01O9tmEdiXQiPu3KH4K00D7w

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_011CcAPRQFTxF1p1cdTJYBm8
O-Max-L2	cf-ray: a0d8cbcb1d5b69c7-LAX server: Photon-Edge set-cookie: _cfuvid=R_9JhbhFc_B2JNuCapgm2C.n2eigtJodP71dZzLQwMw-1781770001.1314807-1.0.1.1-5vxPf6J7uf2uSAIC7xD9fgybSJzhUhOY6UH4DkZ3FxU; 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	1781770001004-4i0u9kn9wioqp req_011CcAP4XPVYcWEmG4EKnp8F a0d8cbcb1d5b69c7-LAX
O-Max	server: nginx x-oneapi-request-id: 20260618081153446096068268d9d6ZyEg4BVe	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-idbody:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	—
R-ultra	cf-ray: a0d8d4ff0fdbfd4f-NRT server: cloudflare x-new-api-version: v0.0.0 x-oneapi-request-id: 202606180812582474707548268d9d6otrtMWx3	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	a0d8d4ff0fdbfd4f-NRT

协议指纹评分（机器自动）

总分 / 100 100 66 60

基线状态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 0.0 /10 ×20

缓存 TTL 一致 10.0 /10 ×15 — 0.0 /10 ×15 0.0 /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 10.0 /10 ×13 — 10.0 /10 ×13 5.3 /10 ×13

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

系统提示纯净 10.0 /10 ×8 — 0.0 /10 ×8 0.0 /10 ×8

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

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

身份(自由) 10.0 /10 ×7 — 10.0 /10 ×7 0.0 /10 ×7

service_tier 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

延迟基线 10.0 /10 ×5 — 3.1 /10 ×5 10.0 /10 ×5

流式投递 10.0 /10 ×5 — 10.0 /10 ×5 6.9 /10 ×5

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

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

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

综合结论

与基线相似度基准（参考）执行失败，不评分 66 v3.24.0 60 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 3642 ms 3750 ms 5524 ms 7154 ms

输出速度 94.3 tok/s 130.6 tok/s 112.3 tok/s 91.3 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

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.3046

$0.2877 (部分)

$0.2077

$0.3816

缓存读取占比 ⓘ 71% 71% (部分) 72% 23%

揭晓

通道

官方基准

服务商 SSSAiCode ↗

通道O-Max-L2

服务商 Novaris ↗

通道O-Max

服务商 FoxCode ↗

通道R-ultra

返回列表