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

2026/06/29 运行日期

claude-sonnet-4-6 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

Aiberm / cc
```
评测失败
```

步骤 / 指标

HongMaCC

u-cc

AIMZ

M-Max

FoxCode

R-ultra

官方基准

基本信息

服务名 Claude Code Claude Code Claude Code Claude Code

模型 claude-sonnet-4-6 claude-sonnet-4-6 claude-sonnet-4-6 claude-sonnet-4-6

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 17:43:27 ~ 17:43:32 (3.5s) 17:43:21 ~ 17:43:26 (4.4s) 17:43:23 ~ 17:43:28 (4.0s) 17:43:29 ~ 17:43:34 (4.0s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 2,110 ms 3,071 ms 2,468 ms 2,822 ms

首文本(SSE delta) 2,111 ms 3,154 ms 2,469 ms 2,824 ms

总耗时 2,111 ms 3,286 ms 3,018 ms 2,910 ms

输入 0 3 96 3

输出 25 5 14 5

缓存创建

5m: 0 / 1h: 0

31,985

5m: 0 / 1h: 31,985

32,920

5m: 32,920 / 1h: 0

32,318

5m: 0 / 1h: 32,318

缓存读取 0 0 0 0

响应

pong

pong

pong

pong

步骤 2: identity text

时间 17:47:24 ~ 17:47:29 (4.8s) 17:46:52 ~ 17:46:57 (4.2s) 17:45:00 ~ 17:45:10 (5.0s) 17:46:40 ~ 17:46:45 (3.4s)

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) 3,666 ms 2,668 ms 2,284 ms 2,037 ms

首文本(SSE delta) 3,667 ms 2,928 ms 2,285 ms 2,038 ms

总耗时 3,667 ms 3,058 ms 4,167 ms 2,395 ms

输入 0 3 49 3

输出 184 22 77 22

缓存创建

5m: 0 / 1h: 0

192

5m: 0 / 1h: 192

567

5m: 567 / 1h: 0

192

5m: 0 / 1h: 192

缓存读取 0 31,985 32,920 32,318

输出速度 burst - 40.9 tok/s -

响应

vendor: anthropic
brand: claude
model: claude-sonnet-4-6

vendor: anthropic
brand: claude
model: claude-sonnet-4-5

vendor: anthropic
brand: claude
model: claude-sonnet-4-6

vendor: anthropic
brand: claude
model: claude-sonnet-4-6

步骤 3: cutoff text

时间 17:48:33 ~ 17:48:38 (3.6s) 17:48:51 ~ 17:49:01 (5.8s) 17:47:46 ~ 17:47:56 (6.1s) 17:48:07 ~ 17:48:12 (3.7s)

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) 2,235 ms 4,323 ms 4,994 ms 2,516 ms

首文本(SSE delta) 2,236 ms 4,483 ms 4,995 ms 2,518 ms

总耗时 2,237 ms 4,650 ms 5,258 ms 2,706 ms

输入 0 3 49 3

输出 39 8 36 8

缓存创建

5m: 0 / 1h: 0

5m: 0 / 1h: 55

400

5m: 400 / 1h: 0

5m: 0 / 1h: 57

缓存读取 0 32,177 33,487 32,510

输出速度 burst - 136.9 tok/s -

响应

2025-08

2025-04

2025-08

2025-08

步骤 4: identity_free text

时间 17:50:55 ~ 17:51:00 (4.1s) 17:50:34 ~ 17:50:44 (5.2s) 17:49:34 ~ 17:49:44 (5.1s) 17:51:22 ~ 17:51:27 (4.5s)

提示词

查看本步 Prompt

用中文回答：你是谁？由谁开发？50 字以内。

HTTP 头(TTFB) 2,888 ms 3,273 ms 3,253 ms 3,039 ms

首文本(SSE delta) 2,889 ms 3,856 ms 3,254 ms 3,041 ms

总耗时 2,889 ms 3,993 ms 4,310 ms 3,542 ms

输入 0 3 95 3

输出 50 38 110 41

缓存创建

5m: 0 / 1h: 0

5m: 0 / 1h: 41

338

5m: 338 / 1h: 0

5m: 0 / 1h: 35

缓存读取 0 32,232 33,887 32,567

输出速度 burst - 104.2 tok/s 81.8 tok/s

响应

我是Claude，由Anthropic开发。

我叫 Claude，是由 Anthropic 公司开发的 AI 助手，属于 Claude 产品系列。

我是 Kiro，一款基于 Anthropic Claude 模型构建的 AI 编程助手，专注于软件工程任务，帮助开发者编写、调试和优化代码。

我是Claude，由Anthropic公司开发的AI助手。Anthropic是一家专注于AI安全研究的美国公司。

步骤 5: knowledge_recall text

时间 17:52:06 ~ 17:52:37 (29.1s) 17:54:14 ~ 17:54:24 (6.9s) 17:53:27 ~ 17:54:03 (30.3s) 17:53:53 ~ 17:54:03 (5.3s)

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) 27,513 ms 1,995 ms 2,430 ms 1,786 ms

首文本(SSE delta) 27,514 ms 4,231 ms 2,431 ms 1,788 ms

总耗时 27,853 ms 4,706 ms 29,373 ms 4,183 ms

输入 0 3 48 3

输出 2,950 198 2,345 205

缓存创建

5m: 0 / 1h: 0

7,700

5m: 0 / 1h: 7,700

9,681

5m: 9,681 / 1h: 0

8,032

5m: 0 / 1h: 8,032

缓存读取 0 0 0 0

输出速度 - - 87.0 tok/s 85.6 tok/s

响应

{"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

{"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

{"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

{"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

步骤 6: digit_count text

时间 17:54:06 ~ 17:54:26 (16.5s) 17:55:54 ~ 17:56:14 (16.2s) 17:55:26 ~ 17:55:56 (29.2s) 17:56:24 ~ 17:56:44 (19.0s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=461184682658364348345047694070184105039966922470106966553146092587000894

HTTP 头(TTFB) 14,961 ms 1,966 ms 2,156 ms 1,678 ms

首文本(SSE delta) 14,962 ms 1,967 ms 2,158 ms 2,079 ms

总耗时 15,333 ms 14,955 ms 28,429 ms 18,030 ms

输入 0 110 87 164

输出 3,787 917 2,653 1,317

缓存创建

5m: 0 / 1h: 0

123

5m: 123 / 1h: 0

5m: 0 / 1h: 0

缓存读取 0 0 0 0

输出速度 burst 70.6 tok/s 101.0 tok/s 82.6 tok/s

响应

{"n":72,"c":[11,4,8]}

```json
{"n":72,"c":[11,4,7]}
```

{"n":72,"c":[11,4,7]}

**Parsing `s` digit by digit:**

`s = 461184682658364348345047694070184105039966922470106966553146092587000894`

**Length:**
Segmenting into groups of 10:
`4611846826` | `5836434834` | `5047694070` |

总计

总输入(含缓存) 0 136,492 144,747 138,208

总输出 7,035 1,188 5,235 1,598

总缓存创建

5m: 0 / 1h: 0

39,973

5m: 0 / 1h: 39,973

44,029

5m: 44,029 / 1h: 0

40,634

5m: 0 / 1h: 40,634

总缓存读取 0 96,394 100,294 97,395

输出速度 burst 70.6 tok/s 92.5 tok/s 82.9 tok/s

总执行时间 1m 1.6s 42.7s 1m 19.6s 39.8s

总墙钟时间

10m 59.1s

含步骤间等待

12m 53.0s

含步骤间等待

12m 33.1s

含步骤间等待

13m 14.9s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

>=$0.1055

Anthropic: Claude Sonnet 4.6 (成本不完整)

$0.1970

Anthropic: Claude Sonnet 4.6

$0.2131

Anthropic: Claude Sonnet 4.6

$0.2061

Anthropic: Claude Sonnet 4.6

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 -

ID 格式 msg_8cb2ecee379befb2e95476b0 msg_018A2m8W6sZNRrEKSQWVZWoR msg_01LVx8KbJjkwJ8fUSavYZDNO msg_01W7vyugKV4dLwC6JFcCAHGv

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
u-cc	server: nginx x-new-api-version: 1.0.17 x-oneapi-request-id: 202606290954074050210058268d9d6rV2ke6XS	id_format: msg_prefix extra_fields: ["stop_details"] has_service_tier: true has_inference_geo: true model_has_date_suffix: false usage_has_cache_fields: true	hdr:x-new-api-versionhdr:x-oneapi-request-idbody:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	—
M-Max	cf-ray: a1340eeffe47b594-AMS server: cloudflare x-new-api-version: v0.0.0 x-oneapi-request-id: 202606290955555985677858268d9d6i69BE4RE	id_format: msg_prefix extra_fields: ["stop_details"] has_service_tier: true has_inference_geo: true model_has_date_suffix: false 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:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	a1340eeffe47b594-AMS
R-ultra	cf-ray: a1340e3d79d2af8b-NRT server: cloudflare x-new-api-version: v0.0.0 x-oneapi-request-id: 202606290955271576380348268d9d6nMRNWbwi	id_format: msg_prefix has_service_tier: true has_inference_geo: true model_has_date_suffix: false 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:usage.cache_fieldsupstream:anthropic-direct-from-body	a1340e3d79d2af8b-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: false usage_has_cache_fields: true	body:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	req_011CcXMKxV8M1Nm4kBpayVxk

协议指纹评分（机器自动）

总分 / 100 57 86 83 100

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

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

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

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

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

知识截止 10.0 /10 ×7 5.0 /10 ×7 10.0 /10 ×7 10.0 /10 ×7

身份(自由) 10.0 /10 ×7 10.0 /10 ×7 0.0 /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 10.0 /10 ×5 10.0 /10 ×5 9.6 /10 ×5 10.0 /10 ×5

流式投递 0.0 /10 ×5 0.0 /10 ×5 9.9 /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

系统提示纯净未计入 5.0 /10 ×8 5.0 /10 ×8 10.0 /10 ×8

综合结论

与基线相似度 57 v3.25.0 86 v3.25.0 83 v3.25.0 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 流式投递 buffer_dump_match
6 自由身份 identity_free_clean
7 停止原因 stop_reason_present
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 缓存连续 cache_continuity_intra
13 知识召回 knowledge_recall_match
14 延迟基线 latency_baseline_match
15 原生 msg-ID anthropic_msg_id_format
16 Sliding 缓存 cache_sliding_correctness
17 结构化身份 identity_structured_match
18 自报身份一致 self_identity_consistency
19 信封自报一致 envelope_self_report_match
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 9015 ms 5775 ms 12426 ms 5628 ms

输出速度 burst 70.6 tok/s 92.5 tok/s 82.9 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 ⓘ

>=$0.1055 (成本不完整)

$0.1970

$0.2131

$0.2061

缓存读取占比 ⓘ - 71% 69% 70%

揭晓

通道

服务商 HongMaCC ↗

通道u-cc

服务商 AIMZ ↗

通道M-Max

服务商 FoxCode ↗

通道R-ultra

官方基准

返回列表