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

2 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

LinkAPI

O-Max

FoxCode

R-super

WorldBase.ai

M-

AnyAIGC

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 09:34:40 ~ 09:34:50 (5.0s) 09:34:34 ~ 09:35:04 (27.4s) 09:34:32 ~ 09:34:37 (3.9s) 09:34:38 ~ 09:34:53 (10.3s) 09:34:36 ~ 09:34:46 (6.4s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 3,554 ms 23,600 ms 2,371 ms 8,504 ms 4,620 ms

首文本(SSE delta) 3,556 ms 23,601 ms 2,415 ms 8,575 ms 4,622 ms

总耗时 3,633 ms 25,919 ms 2,944 ms 8,577 ms 4,733 ms

输入 1,454 1,454 61 1,048 2

输出 4 4 2 1 4

缓存创建

30,222

5m: 0 / 1h: 30,222

30,222

5m: 0 / 1h: 30,222

23,499

5m: 23,499 / 1h: 0

20,322

5m: 0 / 1h: 20,322

36,615

5m: 0 / 1h: 36,615

缓存读取 0 0 0 0 0

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 09:38:18 ~ 09:38:28 (6.7s) 09:36:04 ~ 09:36:44 (38.8s) 09:37:55 ~ 09:38:00 (3.5s) 09:37:27 ~ 09:37:37 (9.3s) 09:37:13 ~ 09:37:23 (5.4s)

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) 5,280 ms 36,754 ms 1,845 ms 7,586 ms 3,322 ms

首文本(SSE delta) 5,281 ms 36,755 ms 1,846 ms 7,738 ms 3,323 ms

总耗时 5,627 ms 37,649 ms 2,766 ms 8,134 ms 3,678 ms

输入 2 2 13 16 2

输出 28 28 58 14 28

缓存创建

1,643

5m: 0 / 1h: 1,643

1,643

5m: 0 / 1h: 1,643

23,745

5m: 23,745 / 1h: 0

1,169

5m: 0 / 1h: 1,169

191

5m: 0 / 1h: 191

缓存读取 30,222 30,222 0 20,322 36,615

输出速度 - - 63.0 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 09:42:07 ~ 09:42:17 (5.3s) 09:39:06 ~ 09:40:06 (58.6s) 09:39:10 ~ 09:39:16 (3.3s) 09:39:12 ~ 09:39:22 (6.1s) 09:39:43 ~ 09:39:53 (7.0s)

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) 3,827 ms 53,881 ms 1,747 ms 3,618 ms 5,306 ms

首文本(SSE delta) 3,828 ms 53,882 ms 1,749 ms 3,619 ms 5,307 ms

总耗时 4,115 ms 57,460 ms 2,104 ms 3,648 ms 5,468 ms

输入 2 2 60 24 2

输出 6 6 11 2 6

缓存创建

5m: 0 / 1h: 64

5m: 0 / 1h: 54

355

5m: 355 / 1h: 0

5m: 0 / 1h: 26

5m: 0 / 1h: 64

缓存读取 31,865 31,865 23,745 21,491 36,806

响应

2026-01

2026-01

2026-01

2026-01

2026-01

步骤 4: identity_free text

时间 09:44:35 ~ 09:44:45 (5.8s) 09:43:50 ~ 09:46:45 (171.9s) 09:43:08 ~ 09:43:13 (4.3s) 09:40:30 ~ 09:40:40 (7.8s) 09:42:05 ~ 09:42:15 (9.3s)

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) 3,815 ms 169,613 ms 1,728 ms 5,793 ms 6,913 ms

首文本(SSE delta) 3,816 ms 169,614 ms 1,729 ms 6,499 ms 6,914 ms

总耗时 4,485 ms 170,861 ms 3,522 ms 6,665 ms 8,073 ms

输入 2 2 16 32 2

输出 45 65 72 34 59

缓存创建

5m: 0 / 1h: 37

5m: 0 / 1h: 31

24,501

5m: 24,501 / 1h: 0

5m: 0 / 1h: 17

5m: 0 / 1h: 39

缓存读取 31,929 31,919 0 21,517 36,870

输出速度 67.3 tok/s 52.1 tok/s 40.2 tok/s - 50.9 tok/s

响应

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

我是 Claude，由 Anthropic 开发的 AI 助手。具体模型是 Claude Opus 4.8，可以帮你处理编程、写作、分析等各类任务。

我是 Kiro，由 Anthropic 开发的 AI 助手，基于 Claude 模型构建。我擅长软件开发、代码编写、分析规划等专业任务。

我是 Claude，由 Anthropic 开发的 AI 助手。在此环境中我以 Kiro 的形式协助你完成软件开发和各类任务。

我是 Claude，由 Anthropic 开发的 AI 助手。当前模型为 Claude Opus 4.8，可协助编程、写作、分析等任务。

步骤 5: knowledge_recall text

时间 09:46:27 ~ 09:46:37 (6.2s) 09:50:38 ~ 09:50:48 (9.8s) 09:46:11 ~ 09:46:36 (23.1s) 09:44:26 ~ 09:44:41 (11.4s) 09:46:13 ~ 09:46:33 (15.0s)

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) 1,417 ms 2,947 ms 1,315 ms 6,573 ms 5,481 ms

首文本(SSE delta) 1,905 ms 6,041 ms 1,319 ms 6,574 ms 11,107 ms

总耗时 4,766 ms 8,815 ms 22,083 ms 9,928 ms 13,299 ms

输入 2 2 63 40 2

输出 356 687 1,793 310 789

缓存创建

4,765

5m: 0 / 1h: 4,765

4,769

5m: 0 / 1h: 4,769

4,684

5m: 4,684 / 1h: 0

3,276

5m: 0 / 1h: 3,276

51,329

5m: 0 / 1h: 51,329

缓存读取 0 0 0 0 0

输出速度 124.4 tok/s burst 86.4 tok/s 92.4 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

{"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 09:48:39 ~ 09:48:44 (3.0s) 09:54:37 ~ 09:54:58 (19.9s) 09:49:05 ~ 09:49:20 (11.6s) 09:48:10 ~ 09:48:25 (12.7s) 09:48:12 ~ 09:48:17 (1.7s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=654625376571329983842773398195860487333665122941695046158476985185230482

HTTP 头(TTFB) 1,713 ms 18,573 ms 1,689 ms 5,386 ms -

首文本(SSE delta) 1,715 ms 18,574 ms 1,690 ms 5,387 ms -

总耗时 2,031 ms 18,963 ms 10,776 ms 11,627 ms -

输入 211 211 54 145 -

输出 19 19 1,180 95 -

缓存创建

5m: 0 / 1h: 0

154

5m: 154 / 1h: 0

5m: 0 / 1h: 0

缓存读取 0 0 0 0 -

输出速度 - - 129.9 tok/s 15.2 tok/s -

响应

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

{"n":72,"c":[3,7,6]}

{"n":72,"c":[3,6,7]}

{"n":72,"c":[3,6,7]}

失败

评测失败

总计

总输入(含缓存)

132,420

132,398

100,950

89,445

198,539

仅前 6 步

总输出

458

809

3,116

456

886

仅前 6 步

总缓存创建

36,731

5m: 0 / 1h: 36,731

36,719

5m: 0 / 1h: 36,719

76,938

5m: 76,938 / 1h: 0

24,810

5m: 0 / 1h: 24,810

88,238

5m: 0 / 1h: 88,238

仅前 6 步

总缓存读取

94,016

94,006

23,745

63,330

110,291

仅前 6 步

输出速度

113.6 tok/s

52.1 tok/s

95.3 tok/s

42.2 tok/s

50.9 tok/s

仅前 6 步

总执行时间

32.1s

5m 26.3s

49.7s

57.6s

44.9s

仅前 6 步

总墙钟时间

14m 3.9s

含步骤间等待

20m 23.5s

含步骤间等待

14m 48.0s

含步骤间等待

13m 47.0s

含步骤间等待

13m 41.0s

含步骤间等待

仅前 6 步

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.2964

Anthropic: Claude Opus 4.8

$0.3051

Anthropic: Claude Opus 4.8

$0.3916

Anthropic: Claude Opus 4.8

$0.2047

Anthropic: Claude Opus 4.8

$0.6288

Anthropic: Claude Opus 4.8 (部分)

通道指纹

来源步骤 digit_count digit_count digit_count digit_count knowledge_recall

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

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

CDN - - cloudflare cloudflare -

ID 格式 msg_01Wbj8vv7uUabrMiZiHfrGDX msg_01UxfsRXDSxYNb1h6uG1A1UM msg_01UOuwh9m5FPhCcU9i2JL1UP msg_24ef60bb4ce44be8aa624733602547c3 msg_01B1ye7uiKoj4Scq1jogagum

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_011CcR2hyS4BVvf2xjqs9aex
O-Max	server: nginx set-cookie: server_name_session=9a42ff809bdc7f49eaa4735d20f97eff; Max-Age=86400; httponly; path=/ x-new-api-version: v1.0.0-rc.11 x-oneapi-request-id: 202606260154389435941228268d9d68WK41DJF	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-super	cf-ray: a1188db36e67e393-NRT server: cloudflare x-new-api-version: v0.0.0 x-oneapi-request-id: 202606260149066347808658268d9d6tSDhi7aW	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	a1188db36e67e393-NRT
M-	cf-ray: a1188c5d6fe1b8f4-AMS server: cloudflare	id_format: msg_prefix has_service_tier: false has_inference_geo: false model_has_date_suffix: false usage_has_cache_fields: true	hdr:server:cloudflarehdr:cf-raybody:id:msg_prefixbody:usage.cache_fieldsupstream:anthropic-direct-from-body	e012cc34-f282-495a-b846-65524f2fca86 a1188c5d6fe1b8f4-AMS
M-	server: nginx	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	—

协议指纹评分（机器自动）

总分 / 100 100 95 59 75

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 0.0 /10 ×20 10.0 /10 ×20 —

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

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 5.3 /10 ×13 10.0 /10 ×13 —

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

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

知识截止 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 0.0 /10 ×7 10.0 /10 ×7 —

service_tier 10.0 /10 ×6 10.0 /10 ×6 10.0 /10 ×6 0.0 /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 0.0 /10 ×5 10.0 /10 ×5 9.8 /10 ×5 —

流式投递 10.0 /10 ×5 10.0 /10 ×5 5.4 /10 ×5 1.4 /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） 95 v3.24.1 59 v3.24.1 75 v3.24.1 执行失败，不评分

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 4110 ms 53278 ms 7366 ms 8097 ms 7050 ms

输出速度 113.6 tok/s 52.1 tok/s 95.3 tok/s 42.2 tok/s 50.9 tok/s

完成步骤 6/6 成功 6/6 成功 6/6 成功 6/6 成功 5/6 成功 · 1 失败

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.2964

$0.3051

$0.3916

$0.2047

$0.6288 (部分)

缓存读取占比 ⓘ 71% 71% 24% 71% 56% (部分)

揭晓

通道

官方基准

服务商 LinkAPI ↗

通道O-Max

服务商 FoxCode ↗

通道R-super

服务商 WorldBase.ai ↗

通道M-

服务商 AnyAIGC ↗

通道M-

返回列表