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

2026/07/02 运行日期

claude-haiku-4-5-20251001 官方基准

1 通道分数 < 70 另含 2 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

SudoCode / M-Max
```
评测失败
```

步骤 / 指标

SSSAiCode

O-Max

官方基准

sucui

O-Max-NL

FoxCode

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 19:53:46 ~ 19:53:51 (3.0s) 19:53:48 ~ 19:53:53 (4.7s) 19:53:44 ~ 19:53:49 (4.7s) 19:53:42 ~ 19:53:47 (3.6s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 1,597 ms 2,969 ms 2,854 ms 2,240 ms

首文本(SSE delta) 1,922 ms 2,971 ms 3,287 ms 2,278 ms

总耗时 2,055 ms 3,662 ms 3,596 ms 2,734 ms

输入 10 10 10 10

输出 42 51 55 42

缓存创建

5,765

5m: 0 / 1h: 5,765

5m: 0 / 1h: 0

13,745

5m: 0 / 1h: 13,745

21,545

5m: 21,545 / 1h: 0

缓存读取 15,418 21,545 0 0

输出速度 - 73.8 tok/s 178.0 tok/s 92.1 tok/s

响应

pong

pong

pong

pong

步骤 2: identity text

时间 19:56:27 ~ 19:56:37 (9.1s) 19:56:40 ~ 19:56:45 (4.3s) 19:55:40 ~ 19:55:50 (5.9s) 19:56:25 ~ 19:56:35 (8.7s)

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。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) 1,178 ms 1,470 ms 2,137 ms 2,598 ms

首文本(SSE delta) 8,091 ms 1,474 ms 2,936 ms 2,599 ms

总耗时 8,384 ms 3,324 ms 3,788 ms 7,988 ms

输入 10 10 10 10

输出 597 207 209 237

缓存创建

235

5m: 0 / 1h: 235

244

5m: 0 / 1h: 244

248

5m: 0 / 1h: 248

191

5m: 191 / 1h: 0

缓存读取 21,183 21,545 13,745 21,545

输出速度 burst 111.9 tok/s 245.3 tok/s 44.0 tok/s

响应

vendor: anthropic
brand: claude
model: claude-3-5-sonnet-20241022

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 19:59:56 ~ 20:00:01 (3.1s) 19:58:47 ~ 19:58:52 (3.6s) 19:59:41 ~ 19:59:46 (3.8s) 20:00:32 ~ 20:00:37 (3.6s)

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) 1,273 ms 2,079 ms 2,283 ms 1,914 ms

首文本(SSE delta) 2,040 ms 2,080 ms 2,683 ms 1,915 ms

总耗时 2,284 ms 2,601 ms 2,740 ms 2,750 ms

输入 10 10 10 10

输出 57 63 65 64

缓存创建

637

5m: 0 / 1h: 637

230

5m: 0 / 1h: 230

251

5m: 0 / 1h: 251

5m: 42 / 1h: 0

缓存读取 21,418 21,789 13,993 21,736

输出速度 233.6 tok/s 120.9 tok/s - 76.6 tok/s

响应

2024-04

2025-02

2025-02

2025-02

步骤 4: identity_free text

时间 20:03:20 ~ 20:03:25 (4.7s) 20:02:42 ~ 20:02:48 (4.0s) 20:01:16 ~ 20:01:27 (6.2s) 20:02:41 ~ 20:02:51 (5.0s)

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) 1,710 ms 1,506 ms 2,380 ms 1,931 ms

首文本(SSE delta) 2,665 ms 1,507 ms 3,089 ms 2,021 ms

总耗时 3,689 ms 2,856 ms 3,440 ms 4,169 ms

输入 10 10 10 10

输出 197 203 117 213

缓存创建

5m: 0 / 1h: 95

105

5m: 0 / 1h: 105

5m: 0 / 1h: 99

5m: 43 / 1h: 0

缓存读取 22,055 22,019 14,244 21,778

输出速度 192.4 tok/s 150.5 tok/s - 99.2 tok/s

响应

我是Claude，由Anthropic公司打造的AI助手。我擅长代码编写、调试、分析等软件工程任务，可以通过对话帮助用户完成各种技术工作。

我是Claude Haiku 4.5，由Anthropic公司开发的AI助手。我可以帮助进行软件工程任务、代码编写、分析和问答等各种工作。

我是Claude Haiku 4.5，由Anthropic公司开发的AI助手。我可以帮助编程、分析、写作等多种任务。

我是Claude，由Anthropic公司开发的AI助手。我可以帮助编程、分析、写作等多种任务，具有较强的推理能力和安全性考量。

步骤 5: knowledge_recall text

时间 20:07:06 ~ 20:07:26 (16.0s) 20:04:49 ~ 20:05:14 (23.0s) 20:05:19 ~ 20:05:39 (15.3s) 20:04:09 ~ 20:04:34 (23.6s)

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) 1,363 ms 2,206 ms 2,863 ms 1,768 ms

首文本(SSE delta) 14,677 ms 2,207 ms 3,347 ms 1,777 ms

总耗时 15,002 ms 21,992 ms 14,283 ms 22,849 ms

输入 10 10 10 10

输出 1,970 2,542 2,174 2,223

缓存创建

9,298

5m: 0 / 1h: 9,298

3,473

5m: 0 / 1h: 3,473

3,871

5m: 0 / 1h: 3,871

8,822

5m: 8,822 / 1h: 0

缓存读取 0 5,805 13,720 0

输出速度 burst 128.5 tok/s 198.8 tok/s 105.5 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

{"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 20:10:56 ~ 20:11:11 (13.9s) 20:09:13 ~ 20:09:33 (18.6s) 20:09:04 ~ 20:09:24 (18.3s) 20:07:54 ~ 20:07:59 (2.8s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=687958500335071977288024161989223412375732714482826417854114785175131145

HTTP 头(TTFB) 880 ms 987 ms 1,733 ms -

首文本(SSE delta) 881 ms 988 ms 2,326 ms -

总耗时 13,186 ms 17,572 ms 17,224 ms -

输入 139 193 10 -

输出 2,494 1,885 2,958 -

缓存创建

5m: 0 / 1h: 0

623

5m: 0 / 1h: 623

缓存读取 0 0 13,714 -

输出速度 202.7 tok/s 113.7 tok/s 198.6 tok/s -

响应

```json
{"n":72,"c":[4,10,4]}
```

```json
{"n":72,"c":[4,10,4]}
```

Looking at the string s="687958500335071977288024161989223412375732714482826417854114785175131145", I'll count the digits:

Let me break this down systematically:

**Length of string:** 72 characters

失败

评测失败

总计

总输入(含缓存)

96,293

96,998

88,313

95,752

仅前 6 步

总输出

5,357

4,951

5,578

2,779

仅前 6 步

总缓存创建

16,030

5m: 0 / 1h: 16,030

4,052

5m: 0 / 1h: 4,052

18,837

5m: 0 / 1h: 18,837

30,643

5m: 30,643 / 1h: 0

仅前 6 步

总缓存读取

80,074

92,703

69,416

65,059

仅前 6 步

输出速度

202.5 tok/s

121.4 tok/s

199.9 tok/s

92.9 tok/s

仅前 6 步

总执行时间

49.6s

58.2s

54.2s

47.4s

仅前 6 步

总墙钟时间

17m 25.1s

含步骤间等待

15m 45.0s

含步骤间等待

15m 40.0s

含步骤间等待

14m 16.9s

含步骤间等待

仅前 6 步

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.0550

Anthropic: Claude Haiku 4.5

$0.0393

Anthropic: Claude Haiku 4.5

$0.0584

Anthropic: Claude Haiku 4.5

$0.0444

Anthropic: Claude Haiku 4.5 (部分)

通道指纹

来源步骤 digit_count digit_count digit_count knowledge_recall

平台 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%)

CDN - - cloudflare cloudflare

ID 格式 msg_01DFHnQjMzhHRnFhtimUoyXw msg_01BPG59U4yLyBhductGnHvsX msg_01G6KHCTzugunfSQpg85Xsox msg_01JH18Zc248LCnxDpFHjrwm2

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-Max	cf-ray: a14d8cdc0fc62b99-LAX server: Photon-Edge	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	1782994257180-z0gd8l5xajzql req_011CcdD1iGhZsqiTYsrxgJVk a14d8cdc0fc62b99-LAX
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_011CcdCt8zWyuEPopV4tNNz8
O-Max-NL	via: 1.1 Caddy cf-ray: a14d8a2159e6dbb3-FRA server: cloudflare x-new-api-version: v0.12.10 x-oneapi-request-id: 202607021209055785684368268d9d6W2Z56dHp	id_format: msg_prefix via_chain: ["1.1 Caddy"] 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-rayhdr:viabody:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:model:date_suffixbody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	a14d8a2159e6dbb3-FRA
cc	cf-ray: a14d82e9ddece3bf-NRT server: cloudflare x-new-api-version: v0.0.0 x-oneapi-request-id: 202607021204102642708918268d9d6jM08mvwF	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	a14d82e9ddece3bf-NRT

协议指纹评分（机器自动）

总分 / 100 81 100 91

基线状态已对照基线已对照基线已对照基线

缓存命中比 10.0 /10 ×20 10.0 /10 ×20 10.0 /10 ×20 —

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

延迟基线 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 —

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

sliding 5m 10.0 /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 —

系统提示纯净 5.0 /10 ×8 10.0 /10 ×8 0.0 /10 ×8 —

身份(结构) 6.7 /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 —

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

知识截止 0.0 /10 ×5 10.0 /10 ×5 10.0 /10 ×5 —

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

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

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

SDK 一致 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 —

综合结论

与基线相似度 81 v3.26.0 基准（参考） 91 v3.26.0 执行失败，不评分

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 自报身份一致 self_identity_consistency
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 7433 ms 8668 ms 7512 ms 8098 ms

输出速度 202.5 tok/s 121.4 tok/s 199.9 tok/s 92.9 tok/s

完成步骤 6/6 成功 6/6 成功 6/6 成功 5/6 成功 · 1 失败

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.0550

$0.0393

$0.0584

$0.0444 (部分)

缓存读取占比 ⓘ 83% 96% 79% 68% (部分)

揭晓

通道

服务商 SSSAiCode ↗

通道O-Max

官方基准

服务商 sucui ↗

通道O-Max-NL

服务商 FoxCode ↗

通道cc

返回列表