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

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

WorldBase.ai / M-
```
评测失败
```

步骤 / 指标

AIMZ

M-Max

FoxCode

R-super

Aiide

O-Max

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 v2.1.160 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 12:50:01 ~ 12:50:11 (5.7s) 12:50:05 ~ 12:50:10 (3.8s) 12:49:59 ~ 12:50:09 (6.5s) 12:50:07 ~ 12:50:12 (3.5s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 2,867 ms 2,493 ms 3,367 ms 2,022 ms

首文本(SSE delta) 3,276 ms 2,494 ms 3,368 ms 2,024 ms

总耗时 3,547 ms 2,856 ms 3,883 ms 2,400 ms

输入 10 33,019 10 10

输出 58 39 54 41

缓存创建

32,785

5m: 0 / 1h: 32,785

5m: 0 / 1h: 0

33,261

5m: 0 / 1h: 33,261

33,122

5m: 0 / 1h: 33,122

缓存读取 0 0 0 0

输出速度 214.0 tok/s 107.7 tok/s 104.9 tok/s 109.0 tok/s

响应

pong

pong

pong

pong

步骤 2: identity text

时间 12:51:41 ~ 12:51:51 (9.6s) 12:51:54 ~ 12:51:59 (4.5s) 12:51:30 ~ 12:51:40 (6.3s) 12:53:23 ~ 12:53:28 (4.6s)

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。请只输出以下 3 行内容（label 小写，值小写）：
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) 2,399 ms 1,836 ms 2,689 ms 2,130 ms

首文本(SSE delta) 7,910 ms 1,837 ms 2,697 ms 2,131 ms

总耗时 8,206 ms 3,832 ms 5,077 ms 3,444 ms

输入 10 33,550 10 10

输出 533 286 272 179

缓存创建

238

5m: 0 / 1h: 238

5m: 0 / 1h: 0

234

5m: 0 / 1h: 234

221

5m: 0 / 1h: 221

缓存读取 32,785 0 33,261 33,122

输出速度 burst 143.4 tok/s 114.3 tok/s 136.3 tok/s

响应

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 12:53:51 ~ 12:54:01 (5.9s) 12:53:33 ~ 12:53:38 (4.0s) 12:54:07 ~ 12:54:17 (6.0s) 12:56:22 ~ 12:56:27 (3.2s)

提示词

查看本步 Prompt

Knowledge cutoff date? Reply with YYYY-MM only.

HTTP 头(TTFB) 4,154 ms 2,506 ms 3,949 ms 1,697 ms

首文本(SSE delta) 4,467 ms 2,507 ms 3,950 ms 1,699 ms

总耗时 4,871 ms 3,168 ms 4,383 ms 2,264 ms

输入 10 34,164 10 10

输出 68 52 71 69

缓存创建

556

5m: 0 / 1h: 556

5m: 0 / 1h: 0

295

5m: 0 / 1h: 295

202

5m: 0 / 1h: 202

缓存读取 33,023 0 33,495 33,343

输出速度 168.3 tok/s 78.7 tok/s 164.0 tok/s 122.1 tok/s

响应

2024-04

2025-02

2025-02

2025-02

步骤 4: identity_free text

时间 12:56:45 ~ 12:56:55 (9.0s) 12:56:26 ~ 12:56:36 (9.3s) 12:56:56 ~ 12:57:06 (7.3s) 12:59:11 ~ 12:59:21 (5.7s)

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) 3,116 ms 2,102 ms 2,980 ms 3,260 ms

首文本(SSE delta) 7,579 ms 2,103 ms 2,986 ms 3,302 ms

总耗时 7,809 ms 8,285 ms 5,190 ms 4,636 ms

输入 10 34,574 10 10

输出 595 1,120 312 215

缓存创建

106

5m: 0 / 1h: 106

5m: 0 / 1h: 0

113

5m: 0 / 1h: 113

103

5m: 0 / 1h: 103

缓存读取 33,579 0 33,790 33,545

输出速度 burst 181.2 tok/s 141.6 tok/s 161.2 tok/s

响应

我是Claude，由Anthropic公司开发的AI助手，可帮助进行编程、分析和问题解答。

我是Claude，由Anthropic公司开发的AI助手。当前版本为Haiku 4.5。我可以协助编程、写作、分析、研究等各类专业工作。

我是Claude，由Anthropic公司开发的AI助手。我可以帮助编程、分析、写作等各种任务。当前使用的是Haiku 4.5模型。

我是Claude，由Anthropic公司开发的AI助手。我可以帮助用户编程、分析、写作等各种任务。

步骤 5: knowledge_recall text

时间 12:58:53 ~ 12:59:08 (12.4s) 12:58:55 ~ 12:59:31 (31.6s) 12:59:02 ~ 12:59:32 (25.0s) 13:01:28 ~ 13:01:48 (18.8s)

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) 1,709 ms 1,538 ms 2,453 ms 1,417 ms

首文本(SSE delta) 11,006 ms 1,540 ms 2,464 ms 1,418 ms

总耗时 11,317 ms 30,743 ms 23,784 ms 17,877 ms

输入 10 11,044 10 10

输出 1,608 5,004 2,000 2,685

缓存创建

8,898

5m: 0 / 1h: 8,898

5m: 0 / 1h: 0

8,694

5m: 0 / 1h: 8,694

8,481

5m: 0 / 1h: 8,481

缓存读取 0 0 0 0

输出速度 burst 171.4 tok/s 93.8 tok/s 163.1 tok/s

响应

```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

步骤 6: digit_count text

时间 13:01:34 ~ 13:01:54 (18.9s) 13:00:41 ~ 13:00:51 (8.9s) 13:02:40 ~ 13:03:30 (45.2s) 13:05:29 ~ 13:05:55 (22.9s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=071782642843244651207531390119986548146003627347385194930526140371662778

HTTP 头(TTFB) 1,422 ms 1,383 ms 1,715 ms 1,732 ms

首文本(SSE delta) 1,423 ms 1,384 ms 1,716 ms 1,739 ms

总耗时 17,702 ms 8,259 ms 44,115 ms 21,978 ms

输入 139 208 193 193

输出 2,404 1,459 3,653 3,412

缓存创建

5m: 0 / 1h: 0

缓存读取 0 0 0 0

输出速度 147.7 tok/s 212.2 tok/s 86.2 tok/s 168.6 tok/s

响应

```json
{"n":72,"c":[7,8,5]}
```

```json
{"n":72,"c":[7,8,5]}
```

```json
{"n": 72, "c": [7, 8, 5]}
```

```json
{"n":72,"c":[7,8,5]}
```

总计

总输入(含缓存) 142,159 146,559 143,386 142,382

总输出 5,266 7,960 6,362 6,601

总缓存创建

42,583

5m: 0 / 1h: 42,583

5m: 0 / 1h: 0

42,597

5m: 0 / 1h: 42,597

42,129

5m: 0 / 1h: 42,129

总缓存读取 99,387 0 100,546 100,010

输出速度 149.2 tok/s 175.8 tok/s 91.9 tok/s 163.9 tok/s

总执行时间 1m 1.5s 1m 2.2s 1m 36.3s 58.8s

总墙钟时间

11m 53.1s

含步骤间等待

10m 46.2s

含步骤间等待

13m 31.1s

含步骤间等待

15m 48.0s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.0897

Anthropic: Claude Haiku 4.5

$0.1864

Anthropic: Claude Haiku 4.5

$0.0954

Anthropic: Claude Haiku 4.5

$0.0959

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

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

CDN cloudflare cloudflare cloudflare -

ID 格式 msg_01QB21Nbr4LyQtSjH57R2odd msg_017D3bSTwkoa6aaobfls2LTg msg_0135b9kwWA9QupEPgw85uZWA msg_012doUgjGfHEr1dd2DSiRfid

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-Max	cf-ray: a0d7bca7fa158ed2-AMS server: cloudflare x-new-api-version: v0.0.0 x-oneapi-request-id: 202606180501354253181908268d9d63ONRQ7V1	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-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	a0d7bca7fa158ed2-AMS
R-super	cf-ray: a0d7bb599b73e3bf-NRT server: cloudflare x-new-api-version: v0.0.0 x-oneapi-request-id: 202606180500422002654178268d9d6f0cZnxQh	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	a0d7bb599b73e3bf-NRT
O-Max	cf-ray: a0d7be43594e6702-AMS server: cloudflare x-new-api-version: 0.12.11 x-oneapi-request-id: 202606180502414147236838268d9d6VPgr3HH0	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-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	a0d7be43594e6702-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_011CcA9FGe9y7h9Nbq9Ua9dN

协议指纹评分（机器自动）

总分 / 100 81 68 97 100

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 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 10.0 /10 ×8 10.0 /10 ×8 10.0 /10 ×8

系统提示纯净 5.0 /10 ×8 10.0 /10 ×8 10.0 /10 ×8 10.0 /10 ×8

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

知识截止 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

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

延迟基线 9.7 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5

Req-ID 透传 0.0 /10 ×4 0.0 /10 ×4 0.0 /10 ×4 10.0 /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 81 v3.22.0 68 v3.22.0 97 v3.22.0 基准（参考）

雷达图例

共 21 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 档位思考量 tier_thinking_volume_match
21 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 8909 ms 9524 ms 14405 ms 8767 ms

输出速度 149.2 tok/s 175.8 tok/s 91.9 tok/s 163.9 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 ⓘ

$0.0897

$0.1864

$0.0954

$0.0959

缓存读取占比 ⓘ 70% 0% 70% 70%

揭晓

通道

服务商 AIMZ ↗

通道M-Max

服务商 FoxCode ↗

通道R-super

服务商 Aiide ↗

通道O-Max

官方基准

返回列表