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

2026/06/20 运行日期

claude-opus-4-8 官方基准

3 通道分数 < 70 与基线相似度

已揭晓揭晓进度

步骤 / 指标

FoxCode

R-super

SSSAiCode

R-Kiro

官方基准

LinkAPI

O-Max

ClaudeZZ

R-kiro

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 00:28:47 ~ 00:28:52 (3.3s) 00:28:41 ~ 00:28:46 (3.6s) 00:28:49 ~ 00:28:54 (4.4s) 00:28:45 ~ 00:28:50 (3.9s) 00:28:43 ~ 00:28:53 (5.4s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 2,018 ms 2,583 ms 3,238 ms 2,430 ms 3,583 ms

首文本(SSE delta) 2,102 ms 2,618 ms 3,240 ms 2,431 ms 3,589 ms

总耗时 2,182 ms 2,620 ms 3,386 ms 2,570 ms 4,178 ms

输入 100 0 1,454 1,454 2

输出 2 4 4 4 4

缓存创建

23,459

5m: 23,459 / 1h: 0

186

5m: 0 / 1h: 186

30,222

5m: 0 / 1h: 30,222

27,912

5m: 0 / 1h: 27,912

缓存读取 0 0 30,036 0 0

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 00:31:00 ~ 00:31:05 (4.1s) 00:30:10 ~ 00:30:20 (8.4s) 00:32:20 ~ 00:32:25 (4.1s) 00:30:52 ~ 00:31:02 (5.6s) 00:32:11 ~ 00:32:21 (6.1s)

提示词

查看本步 Prompt

以下是对你自身的 metadata 探针。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。按以下 3 行回答自己的身份，每行一个 label，不要寒暄不要解释：
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) 1,794 ms 7,079 ms 2,250 ms 3,869 ms 4,414 ms

首文本(SSE delta) 1,795 ms 7,104 ms 2,252 ms 3,870 ms 4,417 ms

总耗时 3,214 ms 7,353 ms 3,116 ms 4,359 ms 5,077 ms

输入 85 0 2 2 2

输出 77 25 28 28 28

缓存创建

231

5m: 231 / 1h: 0

1,656

5m: 0 / 1h: 1,656

1,656

5m: 0 / 1h: 1,656

216

5m: 0 / 1h: 216

缓存读取 23,459 0 30,222 30,222 27,912

输出速度 54.3 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

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

步骤 3: cutoff text

时间 00:33:15 ~ 00:33:21 (3.3s) 00:32:46 ~ 00:32:52 (3.2s) 00:35:48 ~ 00:35:53 (3.2s) 00:32:37 ~ 00:32:42 (4.5s) 00:34:25 ~ 00:34:36 (6.9s)

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) 2,056 ms 2,305 ms 1,998 ms 2,874 ms 5,025 ms

首文本(SSE delta) 2,057 ms 2,337 ms 2,000 ms 2,893 ms 5,647 ms

总耗时 2,609 ms 2,451 ms 2,188 ms 3,308 ms 6,046 ms

输入 38 0 2 2 2

输出 11 6 6 6 78

缓存创建

471

5m: 471 / 1h: 0

5m: 0 / 1h: 54

5m: 0 / 1h: 64

100

5m: 0 / 1h: 100

缓存读取 23,690 0 31,878 31,878 28,128

响应

2026-01

2025-01

2026-01

2026-01

2025-01

步骤 4: identity_free text

时间 00:34:49 ~ 00:34:54 (4.6s) 00:36:03 ~ 00:36:08 (3.2s) 00:38:56 ~ 00:39:06 (7.6s) 00:35:15 ~ 00:35:25 (8.6s) 00:38:15 ~ 00:38:25 (7.1s)

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) 1,904 ms 2,113 ms 5,119 ms 6,762 ms 4,559 ms

首文本(SSE delta) 1,905 ms 2,114 ms 5,121 ms 6,776 ms 4,560 ms

总耗时 3,813 ms 2,196 ms 6,519 ms 7,591 ms 5,909 ms

输入 38 0 2 2 2

输出 89 23 71 45 78

缓存创建

355

5m: 355 / 1h: 0

5m: 0 / 1h: 39

5m: 0 / 1h: 37

126

5m: 0 / 1h: 126

缓存读取 24,161 0 31,932 31,942 28,228

输出速度 46.6 tok/s - 50.8 tok/s 55.2 tok/s 57.8 tok/s

响应

我是 Kiro，一款由 Anthropic 开发的 AI 编程助手，基于 Claude 模型构建，帮助开发者完成编码、分析、规划等软件工程任务。

我是 Claude，一个 AI 助手，由 Anthropic 公司开发。

我是 Claude，由 Anthropic 开发的 AI 助手。我可以帮助处理软件工程、写作、分析等多种任务，致力于提供有帮助、诚实且安全的协助。

我是 Claude，由 Anthropic 开发，属于 Claude 模型系列（当前为 Opus 4.8）。

我是 Claude Code，Anthropic 开发的官方命令行编程助手，基于 Claude Opus 4.8 模型，可帮你阅读代码、修改文件、执行命令、完成各类软件工程任务。

步骤 5: knowledge_recall text

时间 00:38:25 ~ 00:39:00 (31.5s) 00:38:09 ~ 00:38:19 (8.3s) 00:41:29 ~ 00:41:39 (5.3s) 00:38:43 ~ 00:38:53 (8.8s) 00:39:54 ~ 00:40:04 (7.2s)

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) 2,907 ms 4,901 ms 1,516 ms 2,209 ms 3,297 ms

首文本(SSE delta) 2,908 ms 6,213 ms 1,517 ms 5,066 ms 3,299 ms

总耗时 30,575 ms 7,618 ms 4,371 ms 7,332 ms 6,351 ms

输入 86 0 2 2 2

输出 1,332 342 331 680 323

缓存创建

4,703

5m: 4,703 / 1h: 0

4,796

5m: 0 / 1h: 4,796

4,778

5m: 0 / 1h: 4,778

11,257

5m: 0 / 1h: 11,257

缓存读取 0 0 0 0 0

输出速度 48.1 tok/s burst 116.0 tok/s burst 105.8 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

{"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 00:42:38 ~ 00:43:03 (23.6s) 00:40:07 ~ 00:40:17 (9.6s) 00:43:41 ~ 00:43:46 (2.5s) 00:41:50 ~ 00:41:55 (2.8s) 00:44:01 ~ 00:44:16 (14.1s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=928584414724318529717245646419203089149405562550142661261837891642248827

HTTP 头(TTFB) 1,871 ms 4,925 ms 1,179 ms 1,427 ms 3,957 ms

首文本(SSE delta) 1,872 ms 8,873 ms 1,180 ms 1,428 ms 12,478 ms

总耗时 22,974 ms 8,914 ms 1,549 ms 1,874 ms 13,081 ms

输入 76 0 211 211 2

输出 2,214 20 19 19 972

缓存创建

132

5m: 132 / 1h: 0

5m: 0 / 1h: 0

5,090

5m: 0 / 1h: 5,090

缓存读取 0 0 0 0 3,067

输出速度 104.9 tok/s - - - -

响应

{"n":72,"c":[4,5,6]}

{"n":72,"c":[4,5,6]}

{"n":72,"c":[3,5,8]}

{"n":72,"c":[3,5,6]}

{"n":72,"c":[4,5,6]}

总计

总输入(含缓存) 101,084 0 132,472 132,472 132,048

总输出 3,725 420 459 782 1,483

总缓存创建

29,351

5m: 29,351 / 1h: 0

6,731

5m: 0 / 1h: 6,731

36,757

5m: 0 / 1h: 36,757

44,701

5m: 0 / 1h: 44,701

总缓存读取 71,310 0 124,068 94,042 87,335

输出速度 71.3 tok/s burst 94.5 tok/s 55.2 tok/s 91.1 tok/s

总执行时间 1m 10.4s 36.3s 27.2s 34.1s 46.7s

总墙钟时间

14m 16.1s

含步骤间等待

11m 36.0s

含步骤间等待

14m 56.9s

含步骤间等待

13m 10.0s

含步骤间等待

15m 33.0s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.2455

Anthropic: Claude Opus 4.8

>=$0.0105

Anthropic: Claude Opus 4.8 (成本不完整)

$0.1239

Anthropic: Claude Opus 4.8

$0.3047

Anthropic: Claude Opus 4.8

$0.3602

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 one-api (100%) direct (75%) direct (100%) one-api (100%) direct (100%)

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

CDN cloudflare - - - -

ID 格式 msg_01I2TAsqC1T4vXBDh7c4XGhm msg_87cc8bbf1f04449fa39dbef9c91c7020 msg_01HWwpBqeYQY1NtAuyNPX5a6 msg_014NBL5cww9MLgQgqSBJYM3d msg_01AWwHLDvZxvti8QxkgoCJNX

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-super	cf-ray: a0e3fcfb5f57e393-NRT server: cloudflare x-new-api-version: v0.0.0 x-oneapi-request-id: 202606191642393508295408268d9d6eHbGF8Ng	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	a0e3fcfb5f57e393-NRT
R-Kiro	server: Photon-Edge x-amzn-requestid: c8167ddc-2b8c-43ec-a253-b71ddd88311c	id_format: msg_prefix has_service_tier: false has_inference_geo: false model_has_date_suffix: false usage_has_cache_fields: true	body:id:msg_prefixbody:usage.cache_fieldsupstream:x-amzn-requestidupstream:anthropic-direct-from-body	1781887208333-iqam9429kdhm0 c8167ddc-2b8c-43ec-a253-b71ddd88311c
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_011CcCxJAqDtYRMHw9yrTda9
O-Max	server: nginx set-cookie: server_name_session=7d88fca8b1426ca76c58d5aeba6fb224; Max-Age=86400; httponly; path=/ x-new-api-version: v1.0.0-rc.11 x-oneapi-request-id: 202606191641515620621308268d9d6n0EX9A0X	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-idbody:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	—
R-kiro	via: 1.1 google cf-ray: 444a2b3b72afe394bcee01d19a898071-IAD server: nginx	id_format: msg_prefix via_chain: ["1.1 google"] 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	req_444a2b3b72afe394bcee01d19a898071 444a2b3b72afe394bcee01d19a898071-IAD

协议指纹评分（机器自动）

总分 / 100 74 51 98 96 80

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 8.4 /10 ×20 0.0 /10 ×20 10.0 /10 ×20 8.6 /10 ×20 7.5 /10 ×20

缓存 TTL 一致 0.0 /10 ×15 0.0 /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 5.0 /10 ×14 10.0 /10 ×14 10.0 /10 ×14 10.0 /10 ×14

sliding 5m 10.0 /10 ×13 5.0 /10 ×13 10.0 /10 ×13 10.0 /10 ×13 10.0 /10 ×13

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

系统提示纯净 0.0 /10 ×8 未计入 10.0 /10 ×8 10.0 /10 ×8 0.0 /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 0.0 /10 ×7 10.0 /10 ×7 10.0 /10 ×7 0.0 /10 ×7

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

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

inference_geo 10.0 /10 ×5 0.0 /10 ×5 10.0 /10 ×5 10.0 /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 10.0 /10 ×5 8.0 /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 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 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 74 v3.24.0 51 v3.24.0 基准（参考） 96 v3.24.0 80 v3.24.0

雷达图例

共 23 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 自报身份一致 self_identity_consistency
20 信封自报一致 envelope_self_report_match
21 中文遵循 instruction_following_lang
22 档位思考量 tier_thinking_volume_match
23 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 10895 ms 5192 ms 3522 ms 4506 ms 6774 ms

输出速度 71.3 tok/s burst 94.5 tok/s 55.2 tok/s 91.1 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 ⓘ

$0.2455

>=$0.0105 (成本不完整)

$0.1239

$0.3047

$0.3602

缓存读取占比 ⓘ 71% - 94% 71% 66%

揭晓

通道

服务商 FoxCode ↗

通道R-super

服务商 SSSAiCode ↗

通道R-Kiro

官方基准

服务商 LinkAPI ↗

通道O-Max

服务商 ClaudeZZ ↗

通道R-kiro

返回列表