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

2026/06/21 运行日期

claude-opus-4-8 官方基准

1 任务失败执行不完整

已揭晓揭晓进度

步骤 / 指标

FastCode

OmniaKey

O-Max

SSSAiCode

O-Max

官方基准

ByteCatCode

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 11:11:55 ~ 11:12:05 (8.2s) 11:11:57 ~ 11:12:07 (5.8s) 11:11:53 ~ 11:11:58 (3.3s) 11:12:01 ~ 11:12:06 (4.2s) 11:11:59 ~ 11:12:04 (4.7s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 4,203 ms 3,868 ms 1,888 ms 2,167 ms 2,637 ms

首文本(SSE delta) 4,205 ms 4,123 ms 1,950 ms 2,169 ms -

总耗时 5,137 ms 4,325 ms 2,031 ms 2,621 ms 2,828 ms

输入 1,454 1,454 1,428 1,454 2

输出 4 4 4 4 4

缓存创建

30,222

5m: 0 / 1h: 30,222

30,222

5m: 0 / 1h: 30,222

29,764

5m: 0 / 1h: 29,764

30,222

5m: 0 / 1h: 30,222

36,617

5m: 0 / 1h: 36,617

缓存读取 0 0 0 0 0

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 11:14:53 ~ 11:14:58 (4.2s) 11:15:43 ~ 11:15:48 (4.7s) 11:15:41 ~ 11:15:46 (3.0s) 11:14:28 ~ 11:14:33 (3.4s) 11:15:57 ~ 11:16:02 (4.7s)

提示词

查看本步 Prompt

输出 3 行 metadata 描述你自己（当前 API 模型）。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。禁止任何额外字符：
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) 2,822 ms 3,115 ms 1,486 ms 1,757 ms 2,277 ms

首文本(SSE delta) 2,823 ms 3,116 ms 1,919 ms 1,759 ms -

总耗时 3,046 ms 3,525 ms 1,985 ms 2,130 ms 2,738 ms

输入 2 2 2 2 2

输出 28 28 32 28 28

缓存创建

1,643

5m: 0 / 1h: 1,643

1,643

5m: 0 / 1h: 1,643

1,617

5m: 0 / 1h: 1,617

1,643

5m: 0 / 1h: 1,643

191

5m: 0 / 1h: 191

缓存读取 30,222 30,222 29,764 30,222 36,617

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-4-1-20250805

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 11:18:14 ~ 11:18:24 (6.7s) 11:19:35 ~ 11:19:40 (4.5s) 11:18:52 ~ 11:19:02 (5.2s) 11:16:26 ~ 11:16:31 (3.4s) 11:18:16 ~ 11:18:26 (5.2s)

提示词

查看本步 Prompt

Knowledge cutoff date? Reply with YYYY-MM only.

HTTP 头(TTFB) 3,124 ms 3,222 ms 3,704 ms 2,179 ms 3,708 ms

首文本(SSE delta) 3,474 ms 3,223 ms 3,894 ms 2,180 ms -

总耗时 3,969 ms 3,402 ms 3,961 ms 2,360 ms 3,926 ms

输入 2 2 2 2 2

输出 6 6 6 6 6

缓存创建

5m: 0 / 1h: 54

5m: 0 / 1h: 64

5m: 0 / 1h: 58

5m: 0 / 1h: 64

缓存读取 31,865 31,865 31,381 31,865 36,808

响应

2026-01

2026-01

2025-01

2026-01

2026-01

步骤 4: identity_free text

时间 11:22:13 ~ 11:22:23 (8.3s) 11:21:34 ~ 11:21:49 (12.1s) 11:20:55 ~ 11:21:00 (3.7s) 11:19:26 ~ 11:19:31 (4.2s) 11:21:23 ~ 11:21:33 (5.0s)

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) 4,622 ms 9,747 ms 1,746 ms 1,949 ms 2,373 ms

首文本(SSE delta) 4,623 ms 9,753 ms 2,834 ms 1,951 ms -

总耗时 6,928 ms 10,890 ms 2,902 ms 3,240 ms 3,670 ms

输入 2 2 2 2 2

输出 69 61 70 71 61

缓存创建

5m: 0 / 1h: 31

5m: 0 / 1h: 39

5m: 0 / 1h: 35

5m: 0 / 1h: 39

缓存读取 31,919 31,929 31,439 31,929 36,872

输出速度 29.9 tok/s 53.6 tok/s - 55.1 tok/s 47.0 tok/s

响应

我是 Claude，由 Anthropic 开发的 AI 助手。当前模型是 Claude Opus 4.8（API 标识 claude-opus-4-8），用于协助编程等任务。

我是 Claude，由 Anthropic 开发的人工智能助手，模型为 Claude Opus 4.8，可协助完成编程、写作、分析等多种任务。

我是 Claude，由 Anthropic 开发的人工智能助手。我可以帮助你完成软件工程、写作、分析等多种任务，致力于提供有用、诚实且安全的协助。

我是 Claude，由 Anthropic 打造的 AI 助手，当前模型为 Claude Opus 4.8。可在终端、桌面及网页等环境中协助完成软件工程等任务。

我是 Claude，由 Anthropic 开发的 AI 助手，可帮助处理编程、写作、分析等多种任务，致力于提供有用、安全的回答。

步骤 5: knowledge_recall text

时间 11:24:11 ~ 11:24:26 (9.9s) 11:25:33 ~ 11:25:43 (9.7s) 11:22:25 ~ 11:22:30 (4.9s) 11:22:09 ~ 11:22:19 (8.3s) 11:25:12 ~ 11:25:22 (8.8s)

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) 3,670 ms 3,447 ms 1,801 ms 2,058 ms 2,374 ms

首文本(SSE delta) 6,508 ms 6,004 ms 1,803 ms 4,861 ms -

总耗时 8,795 ms 8,413 ms 4,154 ms 7,318 ms 7,762 ms

输入 2 2 2 2 2

输出 683 648 344 690 681

缓存创建

4,773

5m: 0 / 1h: 4,773

4,775

5m: 0 / 1h: 4,775

4,302

5m: 0 / 1h: 4,302

4,789

5m: 0 / 1h: 4,789

43,006

5m: 0 / 1h: 43,006

缓存读取 0 0 0 0 0

输出速度 burst burst 146.3 tok/s burst 126.4 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 11:25:53 ~ 11:25:58 (3.4s) 11:28:39 ~ 11:28:44 (4.8s) 11:25:10 ~ 11:25:15 (1.1s) 11:25:00 ~ 11:25:05 (2.8s) 11:26:24 ~ 11:26:45 (16.1s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=709950312335166528171315489160595677669949727852398709630716595039430076

HTTP 头(TTFB) 1,879 ms 2,288 ms - 1,428 ms 4,866 ms

首文本(SSE delta) 1,895 ms 2,289 ms - 1,430 ms -

总耗时 2,189 ms 3,667 ms - 1,813 ms 14,327 ms

输入 211 211 - 211 1

输出 19 19 - 19 149

缓存创建

5m: 0 / 1h: 0

2,237

5m: 0 / 1h: 2,237

缓存读取 0 0 - 0 0

输出速度 - - - - 15.7 tok/s

响应

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

{"n":72,"c":[6,8,12]}

失败

评测失败

{"n":70,"c":[7,8,9]}

{"n":72,"c":[8,9,11]}

总计

总输入(含缓存)

132,402

132,424

129,800

仅前 6 步

132,442

192,462

总输出

809

766

456

仅前 6 步

818

929

总缓存创建

36,723

5m: 0 / 1h: 36,723

36,735

5m: 0 / 1h: 36,735

35,780

5m: 0 / 1h: 35,780

仅前 6 步

36,753

5m: 0 / 1h: 36,753

82,154

5m: 0 / 1h: 82,154

总缓存读取

94,006

94,016

92,584

仅前 6 步

94,016

110,297

输出速度

29.9 tok/s

53.6 tok/s

146.3 tok/s

仅前 6 步

55.1 tok/s

55.2 tok/s

总执行时间

40.8s

41.6s

21.2s

仅前 6 步

26.4s

44.4s

总墙钟时间

14m 2.9s

含步骤间等待

16m 47.0s

含步骤间等待

13m 22.0s

含步骤间等待

仅前 6 步

13m 3.9s

含步骤间等待

14m 45.2s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.3051

Anthropic: Claude Opus 4.8

$0.3041

Anthropic: Claude Opus 4.8

$0.2885

Anthropic: Claude Opus 4.8 (部分)

$0.3055

Anthropic: Claude Opus 4.8

$0.5919

Anthropic: Claude Opus 4.8

通道指纹

来源步骤 digit_count digit_count knowledge_recall digit_count digit_count

平台 direct (100%) one-api (100%) direct (100%) direct (100%) one-api (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_01AyE6JD299xxBRtYtcS6mRM msg_01N35bbcywZnHetJ4qZmu3P6 msg_016Gefe4vWozzVPiUYyL9S1k msg_01ParjGBeeZhWrwBLQ4VMwPh msg_01pddvZgMFngNedy9sNLnx2c

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
cc	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	—
O-Max	cf-ray: a0efecb08bbafeb0-AMS server: cloudflare x-oneapi-request-id: 202606210328415374062588268d9d6EQsiDLfq	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	a0efecb08bbafeb0-AMS
O-Max	cf-ray: a0efe38d0f2c8626-LAX server: Photon-Edge set-cookie: _cfuvid=VvCLFSr9drgxmRYLvLsgvHFBAniutQPhJJDNmOhIWnA-1782012146.7247033-1.0.1.1-aCl__Ymm6A4sx1WkgbFfQhGXm5RQVaNRKuiTSyZrbjk; 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	1782012146689-ibl0n3ymmjahg req_011CcFgp7mQhpLJxUAJh2MbW a0efe38d0f2c8626-LAX
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_011CcFh1ZTo8rm8FsFwxs4FL
O-Max	server: nginx/1.29.8 x-oneapi-request-id: 202606210326268342435898268d9d6s2popiJJ	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	—

协议指纹评分（机器自动）

总分 / 100 98 98 100 88

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 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

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 — 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

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 — 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 10.0 /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 0.0 /10 ×4 — 10.0 /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 98 v3.24.0 98 v3.24.0 执行失败，不评分基准（参考） 88 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 5011 ms 5704 ms 3007 ms 3247 ms 5875 ms

输出速度 29.9 tok/s 53.6 tok/s 146.3 tok/s 55.1 tok/s 55.2 tok/s

完成步骤 6/6 成功 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 ⓘ

$0.3051

$0.3041

$0.2885 (部分)

$0.3055

$0.5919

缓存读取占比 ⓘ 71% 71% 71% (部分) 71% 57%

揭晓

通道

服务商 FastCode ↗

通道cc

服务商 OmniaKey ↗

通道O-Max

服务商 SSSAiCode ↗

通道O-Max

官方基准

服务商 ByteCatCode ↗

通道O-Max

返回列表