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。

4 通道 · 6 步骤

2026/06/18 运行日期

claude-haiku-4-5-20251001 官方基准

1 通道分数 < 70 另含 1 任务失败

已揭晓揭晓进度

⚠ 1 通道不可测 · 已从主对比表移除点击展开原因

Aiberm / cc
```
评测失败
```

步骤 / 指标

官方基准

SSSAiCode

R-Kiro

ClaudeAPI

O-Bdrk

ClaudeCN

O-Max

基本信息

服务名 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

客户端 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 03:09:32 ~ 03:09:42 (5.3s) 03:09:24 ~ 03:09:29 (4.2s) 03:09:28 ~ 03:09:38 (5.2s) 03:09:26 ~ 03:09:31 (4.9s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 3,457 ms 2,876 ms 3,175 ms 3,179 ms

首文本(SSE delta) 3,459 ms 2,907 ms 3,177 ms 3,181 ms

总耗时 4,036 ms 3,316 ms 3,725 ms 3,623 ms

输入 10 0 10 10

输出 42 53 47 53

缓存创建

33,122

5m: 0 / 1h: 33,122

33,122

5m: 0 / 1h: 33,122

33,122

5m: 0 / 1h: 33,122

缓存读取 0 0 0 0

输出速度 72.8 tok/s 129.6 tok/s 85.8 tok/s 119.9 tok/s

响应

pong

pong

pong

pong

步骤 2: identity text

时间 03:11:49 ~ 03:11:59 (7.4s) 03:12:10 ~ 03:12:15 (4.6s) 03:12:41 ~ 03:12:51 (8.7s) 03:11:22 ~ 03:11:32 (6.7s)

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) 2,712 ms 2,110 ms 3,144 ms 3,017 ms

首文本(SSE delta) 2,713 ms 2,111 ms 3,145 ms 3,018 ms

总耗时 6,451 ms 3,777 ms 7,397 ms 5,531 ms

输入 10 0 10 10

输出 231 222 234 290

缓存创建

236

5m: 0 / 1h: 236

241

5m: 0 / 1h: 241

247

5m: 0 / 1h: 247

缓存读取 33,122 0 33,122 33,122

输出速度 61.8 tok/s 133.3 tok/s 55.0 tok/s 115.4 tok/s

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

步骤 3: cutoff text

时间 03:14:48 ~ 03:14:58 (5.1s) 03:13:47 ~ 03:13:52 (3.9s) 03:14:03 ~ 03:14:13 (5.1s) 03:14:50 ~ 03:15:00 (5.5s)

提示词

查看本步 Prompt

Knowledge cutoff date? Reply with YYYY-MM only.

HTTP 头(TTFB) 2,840 ms 2,564 ms 2,649 ms 3,207 ms

首文本(SSE delta) 2,842 ms 2,566 ms 2,650 ms 3,208 ms

总耗时 3,941 ms 3,066 ms 3,853 ms 4,380 ms

输入 10 0 10 10

输出 65 63 67 72

缓存创建

254

5m: 0 / 1h: 254

274

5m: 0 / 1h: 274

313

5m: 0 / 1h: 313

缓存读取 33,358 0 33,363 33,369

输出速度 59.1 tok/s 126.0 tok/s 55.7 tok/s 61.4 tok/s

响应

2025-02

2025-02

2025-02

2025-02

步骤 4: identity_free text

时间 03:17:08 ~ 03:17:18 (7.7s) 03:17:17 ~ 03:17:27 (6.0s) 03:17:19 ~ 03:17:34 (10.0s) 03:18:26 ~ 03:18:36 (6.2s)

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) 3,039 ms 2,749 ms 2,508 ms 2,977 ms

首文本(SSE delta) 3,041 ms 2,751 ms 2,509 ms 2,978 ms

总耗时 6,684 ms 5,103 ms 8,728 ms 5,019 ms

输入 10 0 10 10

输出 241 200 473 390

缓存创建

103

5m: 0 / 1h: 103

105

5m: 0 / 1h: 105

114

5m: 0 / 1h: 114

缓存读取 33,612 0 33,637 33,682

输出速度 66.2 tok/s 85.0 tok/s 76.1 tok/s 191.1 tok/s

响应

我是Claude Haiku 4.5，由Anthropic公司打造的AI助手。Haiku是Claude系列中轻量级、高效的模型版本，专为快速对话和任务优化设计。

我是 Kiro，由 Anthropic 公司开发的 Claude AI 助手。我基于 Claude Haiku 4.5 模型，是一个交互式开发环境，专门帮助用户处理软件工程任务，包括编写代码、解决 bug、重构代码和代码审查等。

我是Claude，由Anthropic打造的AI助手。可帮助编程、代码分析、问题解答等多种任务。

我是Claude，由Anthropic公司开发的AI助手。我可以帮助用户进行各种编程、写作、分析等任务。我的知识库截至2025年2月。

步骤 5: knowledge_recall text

时间 03:19:53 ~ 03:20:58 (64.4s) 03:20:35 ~ 03:20:50 (14.9s) 03:18:58 ~ 03:19:38 (35.8s) 03:19:45 ~ 03:20:10 (23.4s)

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) 3,059 ms 1,671 ms 1,932 ms 2,359 ms

首文本(SSE delta) 3,061 ms 1,684 ms 1,933 ms 2,360 ms

总耗时 63,378 ms 14,158 ms 34,534 ms 22,113 ms

输入 10 0 10 10

输出 4,300 2,204 2,516 3,042

缓存创建

8,574

5m: 0 / 1h: 8,574

8,833

5m: 0 / 1h: 8,833

8,804

5m: 0 / 1h: 8,804

缓存读取 0 0 0 0

输出速度 71.3 tok/s 176.7 tok/s 77.2 tok/s 154.0 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":"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

```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

步骤 6: digit_count text

时间 03:22:59 ~ 03:23:14 (14.7s) 03:24:27 ~ 03:24:47 (19.4s) 03:22:10 ~ 03:22:40 (27.7s) 03:23:54 ~ 03:24:09 (13.4s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=603634058438454534159767997028699747836490265510329474041119481637997569

HTTP 头(TTFB) 1,100 ms 1,564 ms 1,402 ms 1,783 ms

首文本(SSE delta) 1,102 ms 1,590 ms 1,403 ms 1,784 ms

总耗时 13,714 ms 18,656 ms 26,347 ms 12,215 ms

输入 193 0 193 193

输出 2,203 2,744 1,991 1,893

缓存创建

5m: 0 / 1h: 0

缓存读取 0 0 0 0

输出速度 174.7 tok/s 160.8 tok/s 79.8 tok/s 181.5 tok/s

响应

```json
{"n":72,"c":[6,8,11]}
```

```json
{"n":72,"c":[6,8,11]}
```

```json
{"n":72,"c":[6,8,11]}
```

```json
{"n":72,"c":[6,8,11]}
```

总计

总输入(含缓存) 142,624 0 142,940 143,016

总输出 7,082 5,486 5,328 5,740

总缓存创建

42,289

5m: 0 / 1h: 42,289

42,575

5m: 0 / 1h: 42,575

42,600

5m: 0 / 1h: 42,600

总缓存读取 100,092 0 100,122 100,173

输出速度 86.4 tok/s 159.2 tok/s 76.4 tok/s 157.9 tok/s

总执行时间 1m 44.5s 52.9s 1m 32.6s 1m 0.1s

总墙钟时间

13m 42.0s

含步骤间等待

15m 23.1s

含步骤间等待

13m 12.1s

含步骤间等待

14m 43.0s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.0985

Anthropic: Claude Haiku 4.5

>=$0.0274

Anthropic: Claude Haiku 4.5 (成本不完整)

$0.0901

Anthropic: Claude Haiku 4.5

$0.0922

Anthropic: Claude Haiku 4.5

通道指纹

来源步骤 digit_count digit_count digit_count digit_count

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

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

CDN - - - cloudflare

ID 格式 msg_01RrBm1N3C5rvz52oG8rj8oj msg_d6cede0655bf4860b680028bd1421b07 msg_01VLPr6AjD47ddUBHDEZ3YaN msg_01UUdjP5bqnN14Cu7EPFkNDn

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_011Cc9NpvY999vVJqWhDEm11
R-Kiro	server: Photon-Edge x-amzn-requestid: 24fc3e23-bb8a-43ec-b613-c741b78b37b8	id_format: msg_prefix has_service_tier: false has_inference_geo: false model_has_date_suffix: false usage_has_cache_fields: true	body:id:msg_prefixbody:usage.cache_fieldsupstream:x-amzn-requestidupstream:anthropic-direct-from-body	1781724268769-ugd8nllxezs10 24fc3e23-bb8a-43ec-b613-c741b78b37b8
O-Bdrk	x-new-api-version: v1.0.0-rc.11 x-oneapi-request-id: 202606171922122519484508268d9d6YwXcrSGV	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	cf-ray: a0d46e7b4efcfc46-AMS server: cloudflare x-oneapi-request-id: 202606180323566137615592XoYak9H	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-idhdr:server:cloudflarehdr:cf-raybody:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:model:date_suffixbody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	a0d46e7b4efcfc46-AMS

协议指纹评分（机器自动）

总分 / 100 100 53 98 98

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 0.0 /10 ×20 10.0 /10 ×20 10.0 /10 ×20

缓存 TTL 一致 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

缓存连续性 10.0 /10 ×14 5.0 /10 ×14 10.0 /10 ×14 10.0 /10 ×14

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

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

系统提示纯净 10.0 /10 ×8 未计入 10.0 /10 ×8 10.0 /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 0.0 /10 ×7 10.0 /10 ×7 10.0 /10 ×7

service_tier 10.0 /10 ×6 0.0 /10 ×6 10.0 /10 ×6 10.0 /10 ×6

inference_geo 10.0 /10 ×5 0.0 /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

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

流式投递未计入未计入未计入未计入

综合结论

与基线相似度基准（参考） 53 v3.22.0 98 v3.22.0 98 v3.22.0

雷达图例

共 23 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 自报身份一致 self_identity_consistency
20 信封自报一致 envelope_self_report_match
21 中文遵循 instruction_following_lang
22 档位思考量 tier_thinking_volume_match
23 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 16367 ms 8013 ms 14097 ms 8814 ms

输出速度 86.4 tok/s 159.2 tok/s 76.4 tok/s 157.9 tok/s

完成步骤 6/6 成功 6/6 成功 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 ⓘ

$0.0985

>=$0.0274 (成本不完整)

$0.0901

$0.0922

缓存读取占比 ⓘ 70% - 70% 70%

揭晓

通道

官方基准

服务商 SSSAiCode ↗

通道R-Kiro

服务商 ClaudeAPI ↗

通道O-Bdrk

服务商 ClaudeCN ↗

通道O-Max

返回列表