模型对比 | RPDiag

测试说明点击展开 5 步 quick-probe-v1 详解

同一会话连续 5 步，覆盖 5 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。步与步之间随机延迟 1–4 分钟。第 3 步（cutoff）累计跨度 > 5 分钟保住 sliding cache 检测；全流程总跨度 > 6 分钟容纳新增的第 5 步。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 · 5 步骤

2026/06/16 运行日期

claude-haiku-4-5-20251001 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

官方基准

SSSAiCode

O-Max-L2

FoxCode

R-super

SudoCode

M-Max

ByteCatCode

O-Max

基本信息

服务名 Claude Code Claude Code Claude Code Claude Code Claude Code

模型 claude-haiku-4-5-20251001 claude-haiku-4-5-20251001 claude-haiku-4-5-20251001 claude-haiku-4-5-20251001 claude-haiku-4-5-20251001

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 22:05:06 ~ 22:05:11 (3.2s) 22:05:02 ~ 22:05:07 (3.2s) 22:04:58 ~ 22:05:03 (4.3s) 22:05:04 ~ 22:05:14 (5.0s) 22:05:00 ~ 22:05:05 (4.7s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 1,694 ms 2,022 ms 2,535 ms 3,338 ms 3,303 ms

首文本(SSE delta) 1,695 ms 2,261 ms 2,536 ms 3,340 ms -

总耗时 2,003 ms 2,315 ms 3,415 ms 3,657 ms 3,380 ms

输入 10 10 33,020 10 3

输出 58 41 74 45 5

缓存创建

5m: 0 / 1h: 0

32,860

5m: 0 / 1h: 32,860

5m: 0 / 1h: 0

32,974

5m: 0 / 1h: 32,974

32,948

5m: 0 / 1h: 32,948

缓存读取 33,122 0 0 0 0

输出速度 188.3 tok/s - 84.2 tok/s 142.0 tok/s -

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 22:08:30 ~ 22:08:35 (4.4s) 22:06:08 ~ 22:06:18 (6.6s) 22:07:26 ~ 22:07:36 (5.0s) 22:08:09 ~ 22:08:19 (5.9s) 22:06:57 ~ 22:07:07 (7.8s)

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。自报身份，输出严格遵守以下 3 行格式（不允许 markdown、代码块、额外文本）：
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) 1,585 ms 1,379 ms 1,955 ms 3,295 ms 5,638 ms

首文本(SSE delta) 1,586 ms 5,832 ms 1,965 ms 3,297 ms -

总耗时 3,339 ms 5,885 ms 4,116 ms 4,959 ms 6,753 ms

输入 10 10 33,606 10 3

输出 203 440 256 222 26

缓存创建

252

5m: 0 / 1h: 252

235

5m: 0 / 1h: 235

5m: 0 / 1h: 0

192

5m: 0 / 1h: 192

192

5m: 0 / 1h: 192

缓存读取 33,122 32,860 0 32,974 32,948

输出速度 115.8 tok/s burst 119.0 tok/s 133.6 tok/s -

响应

vendor: anthropic
brand: claude
model: claude-haiku-4-5-20251001

vendor: anthropic
brand: claude
model: claude-3-5-sonnet

vendor: anthropic
brand: claude
model: claude-haiku-4-5-20251001

vendor: anthropic
brand: claude
model: claude-haiku-4-5-20251001

vendor: anthropic
brand: claude
model: claude-haiku-4-5-20251001

步骤 3: cutoff text

时间 22:10:26 ~ 22:10:31 (3.4s) 22:09:07 ~ 22:09:12 (2.5s) 22:09:31 ~ 22:09:36 (4.6s) 22:10:39 ~ 22:10:49 (5.9s) 22:08:36 ~ 22:08:46 (7.3s)

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) 1,838 ms 1,517 ms 2,521 ms 4,054 ms 5,989 ms

首文本(SSE delta) 1,840 ms 1,709 ms 2,522 ms 4,055 ms -

总耗时 2,355 ms 1,819 ms 3,696 ms 4,700 ms 6,206 ms

输入 10 10 34,212 10 3

输出 78 61 62 67 8

缓存创建

245

5m: 0 / 1h: 245

482

5m: 0 / 1h: 482

5m: 0 / 1h: 0

5m: 0 / 1h: 59

缓存读取 33,374 33,095 0 33,166 33,140

输出速度 151.5 tok/s - 52.8 tok/s 103.9 tok/s -

响应

2025-02

2024-04

2025-02

2025-02

2025-02

步骤 4: identity_free text

时间 22:13:22 ~ 22:13:32 (5.0s) 22:10:43 ~ 22:10:48 (4.2s) 22:13:26 ~ 22:13:36 (5.8s) 22:13:58 ~ 22:14:08 (5.7s) 22:09:58 ~ 22:10:08 (7.2s)

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) 1,618 ms 1,144 ms 2,089 ms 3,118 ms 3,887 ms

首文本(SSE delta) 1,619 ms 3,145 ms 2,090 ms 3,119 ms -

总耗时 3,891 ms 3,297 ms 4,810 ms 4,596 ms 5,195 ms

输入 10 10 34,624 10 3

输出 445 310 516 222 80

缓存创建

118

5m: 0 / 1h: 118

5m: 0 / 1h: 95

5m: 0 / 1h: 0

5m: 0 / 1h: 39

5m: 0 / 1h: 43

缓存读取 33,619 33,577 0 33,225 33,199

输出速度 195.9 tok/s - 189.7 tok/s 150.3 tok/s 61.2 tok/s

响应

我是Claude，由Anthropic公司开发的AI助手，属于Claude产品系列。我可以帮助用户进行对话、编程、分析等多种任务。

我是Claude，由Anthropic公司开发的AI助手。我可以帮助用户完成各种软件工程任务，包括写代码、分析文本、解答问题等。

我是Claude Haiku 4.5，由Anthropic公司开发。属于Claude AI产品系列，是一个轻量高效的大语言模型，专为快速响应和低成本推理优化。

我是Claude Haiku 4.5，由Anthropic公司打造的AI助手。我是一个大型语言模型，可以进行对话、分析、编码和问题解答等多种任务。

我是 Claude Code，由 Anthropic 开发的 AI 编程助手。我帮助用户进行软件工程任务，包括代码编写、调试、重构等。我能使用多种工具与用户协作完成复杂的编程工作。

步骤 5: knowledge_recall text

时间 22:15:24 ~ 22:15:49 (21.0s) 22:12:01 ~ 22:12:16 (12.0s) 22:15:31 ~ 22:15:46 (12.3s) 22:17:57 ~ 22:18:32 (32.6s) 22:11:47 ~ 22:11:57 (5.4s)

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) 1,676 ms 2,093 ms 1,400 ms 2,349 ms 2,834 ms

首文本(SSE delta) 1,679 ms 10,873 ms 1,401 ms 2,351 ms -

总耗时 19,986 ms 11,171 ms 11,477 ms 31,305 ms 4,344 ms

输入 10 10 10,462 10 3

输出 2,774 1,482 1,724 4,599 191

缓存创建

3,079

5m: 0 / 1h: 3,079

8,600

5m: 0 / 1h: 8,600

5m: 0 / 1h: 0

8,085

5m: 0 / 1h: 8,085

8,081

5m: 0 / 1h: 8,081

缓存读取 5,721 0 0 0 0

输出速度 151.5 tok/s burst 171.1 tok/s 158.8 tok/s 126.5 tok/s

响应

```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

```json
{"facts":{"super_bowl_lviii":"Kansas City Chiefs","oscars2024_best_picture":"Oppenheimer","ucl_2024":"UNKNOWN","nba_2024":"UNKNOWN","copa_america_2024":"UNKNOWN","euro2024":"UNKNOWN","nobel_pe

```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

```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

```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

总计

总输入(含缓存) 142,702 141,854 145,924 140,764 140,625

总输出 3,558 2,334 2,632 5,155 310

总缓存创建

3,694

5m: 0 / 1h: 3,694

42,272

5m: 0 / 1h: 42,272

5m: 0 / 1h: 0

41,349

5m: 0 / 1h: 41,349

41,323

5m: 0 / 1h: 41,323

总缓存读取 138,958 99,532 0 99,365 99,287

输出速度 153.7 tok/s burst 154.8 tok/s 156.0 tok/s 96.2 tok/s

总执行时间 37.0s 28.5s 31.9s 55.1s 32.6s

总墙钟时间

10m 42.9s

含步骤间等待

7m 13.9s

含步骤间等待

10m 47.8s

含步骤间等待

13m 27.9s

含步骤间等待

6m 57.1s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.0364

Anthropic: Claude Haiku 4.5

$0.0745

Anthropic: Claude Haiku 4.5

$0.1591

Anthropic: Claude Haiku 4.5

$0.0874

Anthropic: Claude Haiku 4.5

$0.0631

Anthropic: Claude Haiku 4.5

通道指纹

来源步骤 knowledge_recall knowledge_recall knowledge_recall knowledge_recall knowledge_recall

平台 direct (100%) direct (100%) one-api (100%) one-api (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_01JyLMAKDy66p7vF7scvpY45 msg_01CDey9xrAe8wqry37Px1nqj msg_01BQkuOxIIvgfx2aPrF4rnZH msg_bdrk_01NXLeDr3V1X8aJHCYopcusi msg_01EkLfRw3hmBTFKU9SnTs7d7

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: true usage_has_cache_fields: true	body:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:model:date_suffixbody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	req_011Cc75Zatnc4qNSp6ktUPA9
O-Max-L2	cf-ray: a0ca683d0f398e58-LAX server: Photon-Edge set-cookie: _cfuvid=MsCcJ_UcCx._Kddh3AVPTscZ9LQMtZMPy62ITRbRpaM-1781619122.7309167-1.0.1.1-c3Bcem8DHCZKn5okNSiMl8WGdL7tVeeIgDx3A6bYLpo; 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: true usage_has_cache_fields: true	body:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:model:date_suffixbody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	1781619122687-5kv2e0y7xabgd req_011Cc75JdcqAcWiYWWRyU9VL a0ca683d0f398e58-LAX
R-super	cf-ray: a0ca6d59bb42e041-NRT server: cloudflare x-new-api-version: v0.0.0 x-oneapi-request-id: 202606161415323124347228268d9d60pLZXsXr	id_format: msg_prefix has_service_tier: true has_inference_geo: true model_has_date_suffix: true 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:model:date_suffixbody:usage.cache_fieldsupstream:anthropic-direct-from-body	a0ca6d59bb42e041-NRT
M-Max	server: nginx x-oneapi-request-id: 2026061622175872552014158Zfx6O3	id_format: msg_prefix has_service_tier: false has_inference_geo: false model_has_date_suffix: true usage_has_cache_fields: true	hdr:x-oneapi-request-idbody:id:msg_prefixbody:model:date_suffixbody:usage.cache_fieldsupstream:anthropic-direct-from-body	req_2d178699839f419e
O-Max	server: nginx/1.29.8 x-oneapi-request-id: 202606161411488429031748268d9d6GjNdSbWt	id_format: msg_prefix extra_fields: ["stop_details"] has_service_tier: true has_inference_geo: true model_has_date_suffix: true usage_has_cache_fields: true	hdr:x-oneapi-request-idbody:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:model:date_suffixbody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	—

协议指纹评分（机器自动）

总分 / 100 100 80 67 81 87

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 7.4 /10 ×20 0.0 /10 ×20 7.6 /10 ×20 7.6 /10 ×20

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

sliding 5m 10.0 /10 ×13 10.0 /10 ×13 5.0 /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 0.0 /10 ×8 10.0 /10 ×8

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

身份(结构) 10.0 /10 ×7 6.7 /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 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 0.0 /10 ×6 10.0 /10 ×6

inference_geo 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 0.0 /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 0.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 未计入

Req-ID 透传 10.0 /10 ×4 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 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

与基线相似度基准（参考） 80 v3.19.0 67 v3.19.0 81 v3.19.0 87 v3.19.0

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 Thinking 体量 thinking_volume_match
14 缓存连续 cache_continuity_intra
15 知识召回 knowledge_recall_match
16 延迟基线 latency_baseline_match
17 原生 msg-ID anthropic_msg_id_format
18 Sliding 缓存 cache_sliding_correctness
19 结构化身份 identity_structured_match
20 自报身份一致 self_identity_consistency
21 中文遵循 instruction_following_lang
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 6315 ms 4897 ms 5503 ms 9843 ms 5176 ms

输出速度 153.7 tok/s burst 154.8 tok/s 156.0 tok/s 96.2 tok/s

完成步骤 5/5 成功 5/5 成功 5/5 成功 5/5 成功 5/5 成功

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.0364

$0.0745

$0.1591

$0.0874

$0.0631

缓存读取占比 ⓘ 97% 70% 0% 71% 71%

揭晓

通道

官方基准

服务商 SSSAiCode ↗

通道O-Max-L2

服务商 FoxCode ↗

通道R-super

服务商 SudoCode ↗

通道M-Max

服务商 ByteCatCode ↗

通道O-Max

返回列表