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 另含 1 任务失败

已揭晓揭晓进度

步骤 / 指标

luminai

M-Kiro/Antigravity

官方基准

TopRouterCN

O-Max

HongMaCC

u-cc

LinkAPI

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 22:18:36 ~ 22:18:46 (7.0s) 22:18:40 ~ 22:18:45 (3.6s) 22:18:32 ~ 22:18:37 (3.9s) 22:18:38 ~ 22:18:43 (3.5s) 22:18:34 ~ 22:18:44 (5.1s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 4,240 ms 2,008 ms 2,965 ms 2,151 ms 3,246 ms

首文本(SSE delta) 4,242 ms 2,009 ms 2,966 ms 2,152 ms 3,263 ms

总耗时 4,431 ms 2,439 ms 3,041 ms 2,152 ms 3,794 ms

输入 1,037 10 1,645 0 10

输出 1 50 56 26 52

缓存创建

2,225

5m: 2,225 / 1h: 0

33,122

5m: 0 / 1h: 33,122

31,487

5m: 0 / 1h: 31,487

5m: 0 / 1h: 0

33,122

5m: 0 / 1h: 33,122

缓存读取 25,593 0 0 0 0

输出速度 - 116.3 tok/s - - 97.9 tok/s

响应

pong

pong

pong

pong

pong

步骤 2: identity text

时间 22:21:38 ~ 22:21:48 (8.2s) 22:20:24 ~ 22:20:39 (4.1s) 22:22:27 ~ 22:22:42 (10.6s) 22:21:57 ~ 22:22:02 (3.9s) 22:22:25 ~ 22:22:30 (2.7s)

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。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) 4,821 ms 1,595 ms 9,661 ms 2,755 ms -

首文本(SSE delta) 4,822 ms 1,598 ms 9,662 ms 2,757 ms -

总耗时 5,944 ms 3,096 ms 9,740 ms 2,757 ms -

输入 185 10 10 0 -

输出 28 166 208 178 -

缓存创建

1,731

5m: 1,731 / 1h: 0

244

5m: 0 / 1h: 244

1,885

5m: 0 / 1h: 1,885

5m: 0 / 1h: 0

缓存读取 27,129 33,122 31,487 0 -

输出速度 - 110.8 tok/s burst burst -

响应

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

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

失败

评测失败

步骤 3: cutoff text

时间 22:23:13 ~ 22:23:24 (5.6s) 22:24:36 ~ 22:24:42 (3.5s) 22:25:43 ~ 22:25:54 (8.9s) 22:25:22 ~ 22:25:27 (3.5s) -

提示词

查看本步 Prompt

Knowledge cutoff date? Reply with YYYY-MM only.

HTTP 头(TTFB) 4,176 ms 1,966 ms 8,087 ms 2,423 ms -

首文本(SSE delta) 4,177 ms 1,968 ms 8,088 ms 2,424 ms -

总耗时 4,192 ms 2,476 ms 8,165 ms 2,424 ms -

输入 18 10 10 0 -

输出 4 65 66 39 -

缓存创建

1,453

5m: 1,453 / 1h: 0

208

5m: 0 / 1h: 208

248

5m: 0 / 1h: 248

5m: 0 / 1h: 0

缓存读取 27,618 33,366 33,372 0 -

输出速度 - 128.0 tok/s - burst -

响应

2025-02

2025-02

2025-02

2025-02

未执行

步骤 4: identity_free text

时间 22:25:02 ~ 22:25:12 (7.0s) 22:28:07 ~ 22:28:12 (4.8s) 22:28:58 ~ 22:29:23 (20.3s) 22:28:48 ~ 22:28:53 (3.3s) -

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) 4,482 ms 1,695 ms 19,451 ms 2,104 ms -

首文本(SSE delta) 5,003 ms 1,696 ms 19,453 ms 2,105 ms -

总耗时 5,068 ms 3,779 ms 19,543 ms 2,105 ms -

输入 38 10 10 0 -

输出 59 409 333 35 -

缓存创建

1,454

5m: 1,454 / 1h: 0

107

5m: 0 / 1h: 107

33,724

5m: 0 / 1h: 33,724

5m: 0 / 1h: 0

缓存读取 27,640 33,574 0 0 -

输出速度 - 196.4 tok/s burst burst -

响应

我叫Claude Haiku 4.5，由Anthropic公司开发。是Anthropic旗下Claude产品家族中的轻量级模型，专为快速、高效的任务而设计。

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

我是Claude Haiku 4.5，由Anthropic公司打造的AI助手。我可以帮助编程、分析、写作等任务。

我是Claude，由Anthropic开发的AI助手。

未执行

步骤 5: knowledge_recall text

时间 22:28:09 ~ 22:28:19 (5.4s) 22:31:20 ~ 22:31:40 (15.4s) 22:33:17 ~ 22:33:32 (12.6s) 22:30:36 ~ 22:30:46 (7.3s) -

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) 3,031 ms 1,689 ms 9,227 ms 6,040 ms -

首文本(SSE delta) 3,305 ms 1,691 ms 9,228 ms 6,041 ms -

总耗时 3,980 ms 14,387 ms 11,891 ms 6,242 ms -

输入 800 10 10 0 -

输出 192 2,109 1,371 1,304 -

缓存创建

334

5m: 334 / 1h: 0

8,708

5m: 0 / 1h: 8,708

8,675

5m: 0 / 1h: 8,675

5m: 0 / 1h: 0

缓存读取 6,358 0 0 0 -

输出速度 - 166.1 tok/s burst - -

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":"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

{"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

总计

总输入(含缓存)

123,613

142,501

142,563

33,132

仅前 2 步

总输出

284

2,799

2,034

1,582

仅前 2 步

总缓存创建

7,197

5m: 7,197 / 1h: 0

42,389

5m: 0 / 1h: 42,389

76,019

5m: 0 / 1h: 76,019

5m: 0 / 1h: 0

33,122

5m: 0 / 1h: 33,122

仅前 2 步

总缓存读取

114,338

100,062

64,859

仅前 2 步

输出速度

162.6 tok/s

burst

97.9 tok/s

仅前 2 步

总执行时间

33.1s

31.4s

56.3s

21.5s

7.8s

仅前 2 步

总墙钟时间

9m 42.6s

含步骤间等待

12m 59.8s

含步骤间等待

14m 59.8s

含步骤间等待

12m 7.8s

含步骤间等待

3m 56.3s

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 ⓘ

$0.0206

Anthropic: Claude Haiku 4.5

$0.0770

Anthropic: Claude Haiku 4.5

$0.1134

Anthropic: Claude Haiku 4.5

>=$0.0079

Anthropic: Claude Haiku 4.5 (成本不完整)

$0.0417

Anthropic: Claude Haiku 4.5 (部分)

通道指纹

来源步骤 knowledge_recall knowledge_recall knowledge_recall knowledge_recall ping

平台 one-api (100%) direct (100%) direct (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_01WK16DfNrdVUxzn1XDsFTZcah msg_01Lcz1xdfWf3di7R53iSKy9R msg_01VR6JHsFQGWTa7phRsFCQLU msg_1a9482bb82507163121557d3 msg_01ABp2s4hFqjKVvwhH8ZJz4n

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
M-Kiro/Antigravity	cf-ray: a0ca7fde7ae6fb99-AMS server: cloudflare x-oneapi-request-id: 202606161428109781453718268d9d6a4Oda5gc	id_format: msg_prefix extra_fields: ["container"] has_service_tier: true has_inference_geo: false model_has_date_suffix: true usage_has_cache_fields: true	hdr:x-oneapi-request-idhdr:server:cloudflarehdr:cf-raybody:id:msg_prefixbody:usage.service_tierbody:model:date_suffixbody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	a0ca7fde7ae6fb99-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: 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_011Cc76n51RZCemKvwT6mUkD
O-Max	server: nginx/1.27.5	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	tr-ew7y4grd-e7l0kkwoo7-078226
u-cc	server: nginx x-new-api-version: 1.0.10 x-oneapi-request-id: 202606161430375954956818268d9d6qyacxXSR	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-new-api-versionhdr: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	—
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: 202606161418363310022958268d9d67gfZmyF3	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-new-api-versionhdr: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 72 97 82 59

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 7.4 /10 ×20 0.0 /10 ×20 0.0 /10 ×20 —

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

sliding 5m 10.0 /10 ×13 10.0 /10 ×13 7.7 /10 ×13 5.0 /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 0.0 /10 ×8 10.0 /10 ×8 10.0 /10 ×8 未计入 —

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 0.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 —

延迟基线 9.4 /10 ×5 10.0 /10 ×5 2.8 /10 ×5 10.0 /10 ×5 —

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 —

流式投递未计入 10.0 /10 ×5 未计入未计入 —

综合结论

与基线相似度 72 v3.19.0 基准（参考） 82 v3.19.0 59 v3.19.0 执行失败，不评分

雷达图例

共 20 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 自由身份 identity_free_clean
6 停止原因 stop_reason_present
7 系统提示纯净 system_prompt_clean
8 Service Tier service_tier_present
9 缓存命中比 cache_hit_ratio_match
10 缓存 TTL 一致 cache_ttl_consistency
11 推理地区 inference_geo_present
12 Thinking 体量 thinking_volume_match
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 中文遵循 instruction_following_lang
20 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 4723 ms 5235 ms 10476 ms 3136 ms 3794 ms

输出速度 - 162.6 tok/s burst burst 97.9 tok/s

完成步骤 5/5 成功 5/5 成功 5/5 成功 5/5 成功 1/5 成功 · 1 失败 · 3 跳过

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.0206

$0.0770

$0.1134

>=$0.0079 (成本不完整)

$0.0417 (部分)

缓存读取占比 ⓘ 92% 70% 45% - 0% (部分)

揭晓

通道

服务商 luminai ↗

通道M-Kiro/Antigravity

官方基准

服务商 TopRouterCN ↗

通道O-Max

服务商 HongMaCC ↗

通道u-cc

服务商 LinkAPI ↗

通道O-Max

返回列表