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

1 通道分数 < 70 与基线相似度

已揭晓揭晓进度

步骤 / 指标

ClaudeZZ

R-kiro

OmniaKey

O-Max

Aiide

O-Max

官方基准

SAIAi

O-Max

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 20:27:12 ~ 20:27:22 (7.5s) 20:27:04 ~ 20:27:14 (5.4s) 20:27:02 ~ 20:27:12 (6.2s) 20:27:14 ~ 20:27:19 (3.8s) 20:27:06 ~ 20:27:11 (2.8s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 6,204 ms 3,426 ms 3,549 ms 2,663 ms 1,619 ms

首文本(SSE delta) 6,206 ms 3,427 ms 3,588 ms 2,665 ms 1,621 ms

总耗时 6,553 ms 3,575 ms 3,840 ms 2,800 ms 1,714 ms

输入 2 1,454 1,454 1,454 1,454

输出 4 4 4 4 4

缓存创建

27,912

5m: 0 / 1h: 27,912

30,222

5m: 0 / 1h: 30,222

30,380

5m: 0 / 1h: 30,380

5m: 0 / 1h: 0

30,222

5m: 0 / 1h: 30,222

缓存读取 0 0 0 30,222 0

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 20:30:51 ~ 20:30:57 (4.5s) 20:28:26 ~ 20:28:31 (4.8s) 20:29:09 ~ 20:29:19 (8.2s) 20:28:23 ~ 20:28:28 (4.3s) 20:30:36 ~ 20:30:41 (3.2s)

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) 3,006 ms 2,814 ms 6,110 ms 2,868 ms 1,901 ms

首文本(SSE delta) 3,007 ms 2,815 ms 6,129 ms 2,870 ms 1,902 ms

总耗时 3,606 ms 3,447 ms 7,088 ms 3,281 ms 2,242 ms

输入 2 2 2 2 2

输出 28 28 28 28 28

缓存创建

216

5m: 0 / 1h: 216

1,656

5m: 0 / 1h: 1,656

1,656

5m: 0 / 1h: 1,656

1,656

5m: 0 / 1h: 1,656

5m: 0 / 1h: 0

缓存读取 27,912 30,222 30,380 30,222 31,878

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 20:33:57 ~ 20:34:07 (7.3s) 20:30:06 ~ 20:30:21 (10.1s) 20:32:51 ~ 20:33:01 (7.9s) 20:29:48 ~ 20:29:53 (4.3s) 20:33:00 ~ 20:33:05 (2.8s)

提示词

查看本步 Prompt

Knowledge cutoff date? Reply with YYYY-MM only.

HTTP 头(TTFB) 5,994 ms 4,215 ms 6,587 ms 3,094 ms 1,696 ms

首文本(SSE delta) 5,995 ms 4,218 ms 6,594 ms 3,096 ms 1,697 ms

总耗时 6,551 ms 4,604 ms 6,811 ms 3,351 ms 1,881 ms

输入 2 2 2 2 2

输出 6 6 6 6 6

缓存创建

5m: 0 / 1h: 89

5m: 0 / 1h: 62

5m: 0 / 1h: 64

5m: 0 / 1h: 54

5m: 0 / 1h: 62

缓存读取 28,128 31,878 32,036 31,878 31,878

响应

2025-03

2026-01

2026-01

2026-01

2026-01

步骤 4: identity_free text

时间 20:37:03 ~ 20:37:13 (6.1s) 20:33:09 ~ 20:34:19 (69.2s) 20:36:35 ~ 20:36:45 (5.4s) 20:31:37 ~ 20:31:42 (4.3s) 20:35:55 ~ 20:36:00 (4.2s)

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) 4,276 ms 4,729 ms 3,388 ms 2,129 ms 2,071 ms

首文本(SSE delta) 4,590 ms 4,730 ms 4,231 ms 2,131 ms 2,072 ms

总耗时 5,382 ms 4,730 ms 4,232 ms 3,275 ms 3,242 ms

输入 2 2 2 2 2

输出 45 61 50 64 66

缓存创建

5m: 0 / 1h: 52

5m: 0 / 1h: 39

5m: 0 / 1h: 37

5m: 0 / 1h: 31

缓存读取 28,217 31,940 32,100 31,932 31,940

输出速度 56.8 tok/s burst burst 55.9 tok/s 56.4 tok/s

响应

我是 Claude，由 Anthropic 开发，属于 Claude Code 命令行工具中运行的 AI 助手。

我是 Claude，由 Anthropic 开发的 AI 助手，基于 Claude Opus 4.8 模型，可协助编程、写作、分析等多种任务。

我是 Claude，Anthropic 开发的 AI 助手，属于 Claude 模型家族（当前为 Opus 4.8）。

我是 Claude，由 Anthropic 开发的 AI 助手。当前模型是 Claude Opus 4.8，可协助完成软件工程、研究分析等各类任务。

我是 Claude,由 Anthropic 开发的人工智能助手。当前模型是 Claude Opus 4.8,可以帮助你完成编程、写作、分析等各类任务。

步骤 5: knowledge_recall text

时间 20:38:17 ~ 20:38:32 (14.2s) 20:35:57 ~ 20:36:12 (11.3s) 20:39:42 ~ 20:39:57 (14.7s) 20:35:35 ~ 20:35:45 (8.7s) 20:39:19 ~ 20:39:29 (9.3s)

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) 6,321 ms 4,191 ms 5,563 ms 1,669 ms 1,721 ms

首文本(SSE delta) 8,515 ms 7,774 ms 7,253 ms 5,449 ms 5,986 ms

总耗时 11,169 ms 10,001 ms 10,819 ms 7,710 ms 8,352 ms

输入 2 2 2 2 2

输出 614 713 638 688 726

缓存创建

11,106

5m: 0 / 1h: 11,106

4,794

5m: 0 / 1h: 4,794

4,783

5m: 0 / 1h: 4,783

4,781

5m: 0 / 1h: 4,781

3,019

5m: 0 / 1h: 3,019

缓存读取 0 0 0 0 1,772

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 20:40:23 ~ 20:40:43 (16.8s) 20:37:41 ~ 20:37:51 (5.2s) 20:43:12 ~ 20:43:22 (7.0s) 20:37:16 ~ 20:37:21 (2.4s) 20:42:56 ~ 20:43:01 (4.5s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=451852742932237904983472925063529647702981306995403957165006421056221176

HTTP 头(TTFB) 3,008 ms 3,293 ms 5,061 ms 1,215 ms 3,187 ms

首文本(SSE delta) 15,075 ms 3,295 ms 5,062 ms 1,217 ms 3,188 ms

总耗时 15,345 ms 4,046 ms 6,035 ms 1,553 ms 3,536 ms

输入 2 211 211 211 211

输出 1,447 19 19 19 19

缓存创建

5,090

5m: 0 / 1h: 5,090

5m: 0 / 1h: 0

缓存读取 3,067 0 0 0 0

输出速度 burst - - - -

响应

{"n":72,"c":[8,7,9]}

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

{"n":72,"c":[6,7,9]}

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

{"n":72,"c":[6,9,11]}

总计

总输入(含缓存) 131,801 132,486 133,109 132,449 132,475

总输出 2,144 831 745 809 849

总缓存创建

44,465

5m: 0 / 1h: 44,465

36,773

5m: 0 / 1h: 36,773

36,920

5m: 0 / 1h: 36,920

6,522

5m: 0 / 1h: 6,522

33,334

5m: 0 / 1h: 33,334

总缓存读取 87,324 94,040 94,516 124,254 97,468

输出速度 56.8 tok/s burst burst 55.9 tok/s 56.4 tok/s

总执行时间 56.5s 1m 46.0s 49.4s 27.7s 26.8s

总墙钟时间

13m 31.0s

含步骤间等待

10m 47.0s

含步骤间等待

16m 20.0s

含步骤间等待

10m 6.9s

含步骤间等待

15m 54.9s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.3752

Anthropic: Claude Opus 4.8

$0.3060

Anthropic: Claude Opus 4.8

$0.3050

Anthropic: Claude Opus 4.8

$0.1315

Anthropic: Claude Opus 4.8

$0.2867

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

ID 格式 msg_01WsV4y7AB6h4yY2jAbsMS9Q msg_01BScfjCfCUKjbcm4RAUXuQf msg_01YYEpyJCfJb11XGXFx1ddCr msg_01KW69efNNWyWhmH1AKQPygW msg_013WTobHdkUs4agKqQW934oB

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	via: 1.1 google cf-ray: 229ce66a2c0ebc6192d23a135e4866d7-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_229ce66a2c0ebc6192d23a135e4866d7 229ce66a2c0ebc6192d23a135e4866d7-IAD
O-Max	cf-ray: a11c43ceab360baa-AMS server: cloudflare x-oneapi-request-id: 202606261237433220382088268d9d6FNKGvBRU	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	a11c43ceab360baa-AMS
O-Max	cf-ray: a11c4be128169701-AMS server: cloudflare x-new-api-version: 0.12.11 x-oneapi-request-id: 202606261243137332627598268d9d67WoeirjQ	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	a11c4be128169701-AMS
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_011CcRtAia3pFuHXatUxMcdS
O-Max	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_011CcRtbq8tXReXYyX7z4wjn

协议指纹评分（机器自动）

总分 / 100 80 98 94 100 100

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 10.0 /10 ×20 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 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 10.0 /10 ×13 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 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 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 0.0 /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 10.0 /10 ×6 10.0 /10 ×6

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

延迟基线 7.4 /10 ×5 10.0 /10 ×5 7.9 /10 ×5 10.0 /10 ×5 10.0 /10 ×5

流式投递 10.0 /10 ×5 10.0 /10 ×5 0.0 /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 10.0 /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 80 v3.25.0 98 v3.25.0 94 v3.25.0 基准（参考） 100 v3.25.0

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 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 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 8101 ms 5067 ms 6471 ms 3662 ms 3495 ms

输出速度 56.8 tok/s burst burst 55.9 tok/s 56.4 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 ⓘ

$0.3752

$0.3060

$0.3050

$0.1315

$0.2867

缓存读取占比 ⓘ 66% 71% 71% 94% 74%

揭晓

通道

服务商 ClaudeZZ ↗

通道R-kiro

服务商 OmniaKey ↗

通道O-Max

服务商 Aiide ↗

通道O-Max

官方基准

服务商 SAIAi ↗

通道O-Max

返回列表