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

2026/06/29 运行日期

claude-opus-4-8 官方基准

2 通道分数 < 70 另含 1 任务失败

已揭晓揭晓进度

步骤 / 指标

LinkAPI

O-Max

官方基准

Novaris

O-Max

ByteCatCode

O-Max

WorldBase.ai

M-

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 13:20:08 ~ 13:20:13 (4.3s) 13:20:10 ~ 13:20:15 (4.1s) 13:20:02 ~ 13:20:07 (4.2s) 13:20:04 ~ 13:21:14 (67.5s) 13:20:06 ~ 13:20:11 (4.2s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 3,121 ms 2,870 ms 3,253 ms 33,700 ms 2,819 ms

首文本(SSE delta) 3,123 ms 2,885 ms 3,254 ms - 2,821 ms

总耗时 3,131 ms 3,032 ms 3,439 ms 33,701 ms 2,830 ms

输入 1,454 1,454 1 1,454 1,042

输出 4 4 2 4 2

缓存创建

30,382

5m: 30,382 / 1h: 0

30,222

5m: 0 / 1h: 30,222

14,880

5m: 0 / 1h: 14,880

30,380

5m: 0 / 1h: 30,380

21,772

5m: 0 / 1h: 21,772

缓存读取 0 0 9,565 0 0

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 13:23:12 ~ 13:23:17 (4.5s) 13:23:28 ~ 13:23:33 (4.2s) 13:21:10 ~ 13:21:15 (4.3s) 13:23:30 ~ 13:23:40 (7.0s) 13:21:33 ~ 13:23:43 (126.7s)

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) 3,067 ms 2,203 ms 2,776 ms 3,793 ms -

首文本(SSE delta) 3,068 ms 2,205 ms 2,778 ms - -

总耗时 3,464 ms 2,591 ms 3,613 ms 4,964 ms -

输入 2 2 1 2 -

输出 28 28 28 28 -

缓存创建

1,634

5m: 1,634 / 1h: 0

1,634

5m: 0 / 1h: 1,634

3,849

5m: 0 / 1h: 3,849

1,634

5m: 0 / 1h: 1,634

缓存读取 30,382 30,222 26,386 30,380 -

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-4-20250514

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

失败

评测失败

步骤 3: cutoff text

时间 13:25:03 ~ 13:25:08 (3.8s) 13:26:32 ~ 13:26:37 (3.7s) 13:23:20 ~ 13:23:25 (4.3s) 13:27:11 ~ 13:27:21 (5.1s) -

提示词

查看本步 Prompt

Knowledge cutoff date? Reply with YYYY-MM only.

HTTP 头(TTFB) 2,588 ms 2,579 ms 2,864 ms 2,845 ms -

首文本(SSE delta) 2,589 ms 2,581 ms 2,942 ms - -

总耗时 2,820 ms 2,764 ms 3,465 ms 3,778 ms -

输入 2 2 1 2 -

输出 6 6 33 6 -

缓存创建

5m: 54 / 1h: 0

5m: 0 / 1h: 62

3,888

5m: 0 / 1h: 3,888

5m: 0 / 1h: 54

缓存读取 32,016 31,856 26,410 32,014 -

输出速度 - - 63.1 tok/s - -

响应

2026-01

2026-01

我的知识没有固定的截止日期，涵盖到最近的时间。

2026-01

未执行

步骤 4: identity_free text

时间 13:28:19 ~ 13:28:24 (4.8s) 13:29:14 ~ 13:29:24 (5.0s) 13:25:07 ~ 13:25:12 (4.3s) 13:31:04 ~ 13:31:14 (7.7s) -

提示词

查看本步 Prompt

用中文说说你的身份：什么产品？由谁打造？50 字以内。

HTTP 头(TTFB) 2,855 ms 3,048 ms 3,443 ms 4,890 ms -

首文本(SSE delta) 2,857 ms 3,049 ms 3,444 ms - -

总耗时 3,845 ms 4,051 ms 3,579 ms 6,617 ms -

输入 2 2 1 2 -

输出 60 55 61 64 -

缓存创建

5m: 35 / 1h: 0

5m: 0 / 1h: 39

3,928

5m: 0 / 1h: 3,928

5m: 0 / 1h: 31

缓存读取 32,070 31,918 26,437 32,068 -

输出速度 60.7 tok/s 54.9 tok/s - 37.1 tok/s -

响应

我是 Claude Code，Anthropic 打造的官方命令行编程助手，底层模型为 Claude Opus 4.8，帮助你完成软件工程任务。

我是 Claude，由 Anthropic 开发的人工智能助手，可以帮你写代码、分析问题、解答疑惑和处理各类任务。

我是 Claude Code，Anthropic 的官方命令行工具。底层是 Anthropic 开发的 Claude 模型，在终端中为你提供帮助。

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

未执行

步骤 5: knowledge_recall text

时间 13:29:35 ~ 13:29:45 (6.0s) 13:32:40 ~ 13:32:50 (9.3s) 13:28:54 ~ 13:28:59 (4.9s) 13:34:36 ~ 13:35:11 (30.8s) -

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) 2,100 ms 1,887 ms 2,234 ms 27,359 ms -

首文本(SSE delta) 2,101 ms 5,661 ms 2,236 ms - -

总耗时 5,024 ms 8,350 ms 4,158 ms 29,697 ms -

输入 2 2 1 1 -

输出 327 701 282 565 -

缓存创建

6,754

5m: 6,754 / 1h: 0

4,766

5m: 0 / 1h: 4,766

2,554

5m: 0 / 1h: 2,554

3,484

5m: 0 / 1h: 3,484

缓存读取 0 0 1,639 0 -

输出速度 111.9 tok/s burst 146.7 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

```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

未执行

步骤 6: digit_count text

时间 13:31:33 ~ 13:31:43 (7.6s) 13:35:28 ~ 13:35:33 (2.6s) 13:31:39 ~ 13:32:24 (41.3s) 13:38:31 ~ 13:38:41 (7.8s) -

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=971733284031427575921588231548009344837980303623072784313553117036554843

HTTP 头(TTFB) 2,276 ms 1,265 ms 2,401 ms 5,745 ms -

首文本(SSE delta) 3,706 ms 1,267 ms 2,402 ms - -

总耗时 6,647 ms 1,674 ms 2,930 ms 6,719 ms -

输入 2 211 1 211 -

输出 510 19 18 19 -

缓存创建

2,205

5m: 2,205 / 1h: 0

5m: 0 / 1h: 0

244

5m: 348 / 1h: 0

5m: 0 / 1h: 0

缓存读取 0 0 156 0 -

输出速度 burst - - - -

响应

{"n":72,"c":[7,8,4]}

{"n":72,"c":[5,7,4]}

{"n":78,"c":[6,10,6]}

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

未执行

总计

总输入(含缓存)

136,996

132,392

119,942

131,717

22,814

仅前 2 步

总输出

935

813

424

686

仅前 2 步

总缓存创建

41,064

5m: 41,064 / 1h: 0

36,723

5m: 0 / 1h: 36,723

29,343

5m: 348 / 1h: 29,099

35,583

5m: 0 / 1h: 35,583

21,772

5m: 0 / 1h: 21,772

仅前 2 步

总缓存读取

94,468

93,996

90,593

94,462

仅前 2 步

输出速度

99.0 tok/s

54.9 tok/s

128.8 tok/s

37.1 tok/s

仅前 2 步

总执行时间

31.0s

29.0s

1m 3.3s

2m 5.8s

2m 10.9s

仅前 2 步

总墙钟时间

11m 34.9s

含步骤间等待

15m 23.0s

含步骤间等待

12m 22.0s

含步骤间等待

18m 37.0s

含步骤间等待

3m 37.4s

含步骤间等待

仅前 2 步

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.2383

Anthropic: Claude Opus 4.8

$0.3052

Anthropic: Claude Opus 4.8

$0.2392

Anthropic: Claude Opus 4.8

$0.2951

Anthropic: Claude Opus 4.8

$0.1413

Anthropic: Claude Opus 4.8 (部分)

通道指纹

来源步骤 digit_count digit_count digit_count digit_count ping

平台 one-api (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 (100%) Anthropic 直连 (100%)

CDN - - - - cloudflare

ID 格式 msg_01XK28anjjVQMZYGeyvZGxAT msg_01LATHkcyXHNdrK4trJ2a96e msg_019Cktw8hQ1fqgSYYoer2Phx msg_01J3MkxKLqndLpaZNw7mNTZw msg_O1e5MozFlDyCwlL4gc6mS6lF

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	server: nginx set-cookie: server_name_session=9a42ff809bdc7f49eaa4735d20f97eff; Max-Age=86400; httponly; path=/ x-new-api-version: v1.0.0-rc.15 x-oneapi-request-id: 202606290531342491320818268d9d6MNVNqqnb	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	—
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_011CcX1RtdY6fvoViMxn8SVs
O-Max	server: nginx x-oneapi-request-id: 202606290532176533838738268d9d6UErlwr6e	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	—
O-Max	server: nginx/1.29.8 x-oneapi-request-id: 202606290538324286094038268d9d6d7vxZaK0	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	—
M-	cf-ray: a1327af0eb7b6d07-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	7e727eed-ee0c-45d6-b33d-f8d9daef83e4 a1327af0eb7b6d07-AMS

协议指纹评分（机器自动）

总分 / 100 81 100 81 94

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 0.0 /10 ×15 10.0 /10 ×15 10.0 /10 ×15 7.0 /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 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 —

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 10.0 /10 ×8 0.0 /10 ×8 10.0 /10 ×8 —

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

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

延迟基线 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 8.5 /10 ×5 —

流式投递 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 未计入 —

Req-ID 透传 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 —

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 81 v3.25.0 基准（参考） 81 v3.25.0 94 v3.25.0 执行失败，不评分

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 4155 ms 3744 ms 3531 ms 14246 ms 2830 ms

输出速度 99.0 tok/s 54.9 tok/s 128.8 tok/s 37.1 tok/s -

完成步骤 6/6 成功 6/6 成功 6/6 成功 6/6 成功 1/6 成功 · 1 失败 · 4 跳过

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.2383

$0.3052

$0.2392

$0.2951

$0.1413 (部分)

缓存读取占比 ⓘ 69% 71% 76% 72% 0% (部分)

揭晓

通道

服务商 LinkAPI ↗

通道O-Max

官方基准

服务商 Novaris ↗

通道O-Max

服务商 ByteCatCode ↗

通道O-Max

服务商 WorldBase.ai ↗

通道M-

返回列表