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

2026/06/30 运行日期

claude-haiku-4-5-20251001 官方基准

2 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

FastCode / cc
```
评测失败
```

步骤 / 指标

ClaudeAPI

O-Bdrk

HongMaCC

u-cc

FoxCode

官方基准

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 v2.1.195 claude-cli v2.1.195 claude-cli v2.1.195 claude-cli v2.1.195

偏差标红基于官方基准

步骤 1: ping text

时间 19:53:06 ~ 19:53:31 (23.9s) 19:53:08 ~ 19:53:13 (3.3s) 19:53:02 ~ 19:53:07 (3.3s) 19:53:10 ~ 19:53:15 (3.1s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 17,383 ms 2,089 ms 2,012 ms 1,596 ms

首文本(SSE delta) 17,384 ms 2,090 ms 2,017 ms 1,598 ms

总耗时 20,838 ms 2,090 ms 2,518 ms 2,085 ms

输入 22,135 0 10 10

输出 329 25 53 52

缓存创建

5m: 0 / 1h: 0

6,166

5m: 6,166 / 1h: 0

6,166

5m: 0 / 1h: 6,166

缓存读取 0 0 15,379 15,379

输出速度 95.3 tok/s - 105.8 tok/s 106.8 tok/s

响应

pong

pong

pong

pong

步骤 2: identity text

时间 19:56:48 ~ 19:57:18 (25.7s) 19:56:29 ~ 19:56:39 (5.1s) 19:56:26 ~ 19:56:31 (4.2s) 19:55:00 ~ 19:55:10 (6.3s)

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) 21,538 ms 3,902 ms 1,632 ms 2,847 ms

首文本(SSE delta) 21,539 ms 3,903 ms 1,633 ms 2,849 ms

总耗时 24,135 ms 3,904 ms 3,444 ms 5,261 ms

输入 22,339 0 10 10

输出 687 226 177 203

缓存创建

5m: 0 / 1h: 0

191

5m: 191 / 1h: 0

245

5m: 0 / 1h: 245

缓存读取 0 0 21,545 21,545

输出速度 burst burst 97.7 tok/s 84.2 tok/s

响应

I'm Claude, an AI assistant made by Anthropic.

I cannot provide the information in the format you've requested. That request is asking me to output internal identity information, which I don't disclo

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 20:00:17 ~ 20:00:37 (15.8s) 20:00:26 ~ 20:00:36 (6.6s) 19:59:26 ~ 19:59:31 (2.9s) 19:56:17 ~ 19:56:22 (3.4s)

提示词

查看本步 Prompt

你掌握的世界知识截止于何时？YYYY-MM 格式作答，不解释。

HTTP 头(TTFB) 11,774 ms 5,548 ms 1,455 ms 1,599 ms

首文本(SSE delta) 11,775 ms 5,550 ms 1,456 ms 1,601 ms

总耗时 12,798 ms 5,551 ms 2,083 ms 2,304 ms

输入 22,502 0 10 10

输出 163 43 67 61

缓存创建

5m: 0 / 1h: 0

5m: 42 / 1h: 0

226

5m: 0 / 1h: 226

缓存读取 0 0 21,736 21,790

输出速度 159.3 tok/s burst 106.9 tok/s 86.8 tok/s

响应

2025-01

2025-02

2025-02

2025-02

步骤 4: identity_free text

时间 20:02:21 ~ 20:02:31 (8.9s) 20:02:39 ~ 20:02:44 (3.1s) 20:01:46 ~ 20:01:51 (3.9s) 19:58:02 ~ 19:58:12 (9.5s)

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) 5,609 ms 1,953 ms 1,251 ms 1,461 ms

首文本(SSE delta) 5,610 ms 1,954 ms 1,254 ms 1,462 ms

总耗时 6,443 ms 1,954 ms 3,135 ms 8,553 ms

输入 22,539 0 10 10

输出 99 38 208 727

缓存创建

5m: 0 / 1h: 0

5m: 41 / 1h: 0

103

5m: 0 / 1h: 103

缓存读取 0 0 21,778 22,016

输出速度 118.8 tok/s burst 110.6 tok/s 102.5 tok/s

响应

我是Claude，由Anthropic公司开发的AI助手。

我是Claude，由Anthropic打造的AI助手。

我是Claude，具体是Haiku 4.5版本（claude-haiku-4-5-20251001），属于Claude产品系列，由Anthropic公司开发的大语言模型助手。

我是Claude Haiku 4.5，由Anthropic开发的AI助手。能帮助用户进行代码编写、文件编辑、代码审查等软件工程任务。

步骤 5: knowledge_recall text

时间 20:03:56 ~ 20:04:11 (12.3s) 20:05:19 ~ 20:05:35 (14.9s) 20:03:58 ~ 20:04:13 (13.8s) 20:00:35 ~ 20:01:15 (36.5s)

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) 10,224 ms 13,266 ms 1,509 ms 2,726 ms

首文本(SSE delta) 10,226 ms 13,267 ms 1,511 ms 2,727 ms

总耗时 11,135 ms 13,781 ms 13,024 ms 35,494 ms

输入 10,263 0 10 10

输出 542 2,119 1,802 3,949

缓存创建

5m: 0 / 1h: 0

8,822

5m: 8,822 / 1h: 0

9,797

5m: 0 / 1h: 9,797

缓存读取 0 0 0 0

输出速度 burst - 156.5 tok/s 120.5 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

{"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

```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 20:05:30 ~ 20:06:05 (32.3s) 20:07:09 ~ 20:07:34 (22.2s) 20:07:24 ~ 20:07:59 (31.6s) 20:05:01 ~ 20:05:21 (18.5s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=308955521172900359134821606860505193795772565133641055796222578485285398

HTTP 头(TTFB) 2,206 ms 20,828 ms 1,747 ms 1,171 ms

首文本(SSE delta) 2,207 ms 20,829 ms 1,748 ms 1,172 ms

总耗时 31,095 ms 21,115 ms 30,855 ms 17,536 ms

输入 193 0 193 193

输出 3,614 2,352 3,583 1,906

缓存创建

5m: 0 / 1h: 0

缓存读取 0 0 0 0

输出速度 125.1 tok/s burst 123.1 tok/s 116.5 tok/s

响应

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

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

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

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

总计

总输入(含缓存) 99,971 0 95,943 97,510

总输出 5,434 4,803 5,890 6,898

总缓存创建

5m: 0 / 1h: 0

15,262

5m: 15,262 / 1h: 0

16,537

5m: 0 / 1h: 16,537

总缓存读取 0 0 80,438 80,730

输出速度 123.0 tok/s burst 129.6 tok/s 115.3 tok/s

总执行时间 1m 58.9s 55.3s 59.7s 1m 17.3s

总墙钟时间

12m 59.3s

含步骤间等待

14m 26.4s

含步骤间等待

14m 57.0s

含步骤间等待

12m 11.1s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.1271

Anthropic: Claude Haiku 4.5

>=$0.0240

Anthropic: Claude Haiku 4.5 (成本不完整)

$0.0497

Anthropic: Claude Haiku 4.5

$0.0635

Anthropic: Claude Haiku 4.5

通道指纹

来源步骤 digit_count digit_count digit_count digit_count

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

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

CDN - - cloudflare -

ID 格式 msg_01SnRJ5XkGgngWS7JxUYMxoX msg_8810e593ac2452f23e02d335 msg_012RzocqeRsstMa4JpH6EVYM msg_01Bk8XE5B4KduN5FdKsWJPHC

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
O-Bdrk	x-new-api-version: v1.0.0-rc.11 x-oneapi-request-id: 202606301205315743383368268d9d6ZcdJaCVp	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	—
u-cc	server: nginx x-new-api-version: 1.0.17 x-oneapi-request-id: 202606301207105396429918268d9d6E2LcYzP1	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	—
cc	cf-ray: a13d0cebd966e393-NRT server: cloudflare set-cookie: _cfuvid=TGtvHRDfRT7iWNEqQiuvewA0Ib.UWtr3f1DD150uVRM-1782821245.1753228-1.0.1.1-XCl0lBV0MC93kRI7weSbps7SbVwoaMf5Ocuyv8N77cU; 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	hdr: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	req_011CcZR8F8dL7DQCUFcHSmgM a13d0cebd966e393-NRT
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_011CcZQwiMRpnqNfFcoaLRrP

协议指纹评分（机器自动）

总分 / 100 53 63 91 100

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

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

延迟基线 0.0 /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 5.0 /10 ×14 5.0 /10 ×14 10.0 /10 ×14 10.0 /10 ×14

sliding 5m 5.0 /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 5.0 /10 ×8 未计入 10.0 /10 ×8 10.0 /10 ×8

身份(结构) 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

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

知识截止 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5

inference_geo 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5

stop_reason 10.0 /10 ×3 10.0 /10 ×3 10.0 /10 ×3 10.0 /10 ×3

Req-ID 透传 0.0 /10 ×2 0.0 /10 ×2 10.0 /10 ×2 10.0 /10 ×2

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 53 v3.26.0 63 v3.26.0 91 v3.26.0 基准（参考）

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 档位思考量 tier_thinking_volume_match
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 17741 ms 8066 ms 9177 ms 11872 ms

输出速度 123.0 tok/s burst 129.6 tok/s 115.3 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 ⓘ

$0.1271

>=$0.0240 (成本不完整)

$0.0497

$0.0635

缓存读取占比 ⓘ 0% - 84% 83%

揭晓

通道

服务商 ClaudeAPI ↗

通道O-Bdrk

服务商 HongMaCC ↗

通道u-cc

服务商 FoxCode ↗

通道cc

官方基准

返回列表