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

2026/06/28 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

DBAI / M-
```
评测失败
```

步骤 / 指标

官方基准

Novaris

O-Max

HongMaCC

u-cc

SudoCode

M-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 00:52:43 ~ 00:52:48 (3.2s) 00:52:39 ~ 00:52:44 (4.4s) 00:52:35 ~ 00:52:40 (4.3s) 00:52:41 ~ 00:52:46 (4.3s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 1,897 ms 2,551 ms 2,999 ms 2,779 ms

首文本(SSE delta) 1,899 ms 2,553 ms 3,001 ms 2,780 ms

总耗时 2,150 ms 3,122 ms 3,001 ms 3,338 ms

输入 10 10 0 10

输出 38 48 36 62

缓存创建

33,122

5m: 0 / 1h: 33,122

33,122

5m: 0 / 1h: 33,122

5m: 0 / 1h: 0

33,122

5m: 0 / 1h: 33,122

缓存读取 0 0 0 0

输出速度 151.4 tok/s 84.4 tok/s burst 111.1 tok/s

响应

pong

pong

pong

pong

步骤 2: identity text

时间 00:54:53 ~ 00:54:58 (4.1s) 00:54:11 ~ 00:54:21 (5.3s) 00:56:35 ~ 00:56:40 (4.6s) 00:54:27 ~ 00:54:37 (5.4s)

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。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) 1,617 ms 2,150 ms 3,302 ms 2,612 ms

首文本(SSE delta) 1,619 ms 2,151 ms 3,304 ms 2,613 ms

总耗时 3,117 ms 4,499 ms 3,304 ms 4,160 ms

输入 10 10 0 10

输出 179 220 162 186

缓存创建

218

5m: 0 / 1h: 218

228

5m: 0 / 1h: 228

5m: 0 / 1h: 0

242

5m: 0 / 1h: 242

缓存读取 33,122 33,122 0 33,122

输出速度 119.5 tok/s 93.7 tok/s burst 120.2 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 00:58:57 ~ 00:59:02 (3.0s) 00:55:30 ~ 00:55:35 (3.5s) 00:59:08 ~ 00:59:13 (3.9s) 00:57:52 ~ 00:57:57 (4.7s)

提示词

查看本步 Prompt

Knowledge cutoff date? Reply with YYYY-MM only.

HTTP 头(TTFB) 1,674 ms 2,000 ms 2,599 ms 2,765 ms

首文本(SSE delta) 1,675 ms 2,001 ms 2,601 ms 2,766 ms

总耗时 2,046 ms 2,777 ms 2,601 ms 3,345 ms

输入 10 10 0 10

输出 62 72 33 61

缓存创建

202

5m: 0 / 1h: 202

260

5m: 0 / 1h: 260

5m: 0 / 1h: 0

209

5m: 0 / 1h: 209

缓存读取 33,340 33,350 0 33,364

输出速度 167.1 tok/s 92.8 tok/s burst 105.4 tok/s

响应

2025-02

2025-02

2025-02

2025-02

步骤 4: identity_free text

时间 01:02:39 ~ 01:02:54 (10.8s) 00:57:36 ~ 00:57:46 (8.0s) 01:02:14 ~ 01:02:24 (8.7s) 01:00:19 ~ 01:00:29 (7.0s)

提示词

查看本步 Prompt

请用中文介绍你自己是什么产品或助手，由谁开发，50 字以内。

HTTP 头(TTFB) 1,659 ms 2,951 ms 7,610 ms 2,808 ms

首文本(SSE delta) 1,661 ms 2,959 ms 7,611 ms 2,809 ms

总耗时 9,869 ms 7,242 ms 7,611 ms 5,902 ms

输入 10 10 0 10

输出 1,225 492 53 435

缓存创建

104

5m: 0 / 1h: 104

106

5m: 0 / 1h: 106

5m: 0 / 1h: 0

103

5m: 0 / 1h: 103

缓存读取 33,542 33,610 0 33,573

输出速度 149.2 tok/s 114.9 tok/s burst 140.6 tok/s

响应

我是Claude Haiku，由Anthropic公司开发的AI助手，专门协助软件工程任务，包括编程、调试、代码审查和文件操作等。

我是Claude，由Anthropic开发的AI助手。可帮助编程、分析、写作等任务。

我是Claude，由Anthropic开发的AI助手。

我是Claude，由Anthropic公司开发的AI助手。我帮助用户进行软件工程任务，如编写、调试、优化和重构代码。我可以理解代码、提供技术建议和解决问题。

步骤 5: knowledge_recall text

时间 01:04:59 ~ 01:05:19 (16.2s) 01:00:23 ~ 01:00:53 (29.3s) 01:04:55 ~ 01:05:05 (8.2s) 01:03:02 ~ 01:03:17 (13.6s)

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) 2,313 ms 1,545 ms 6,691 ms 1,994 ms

首文本(SSE delta) 2,315 ms 1,547 ms 6,692 ms 1,995 ms

总耗时 15,079 ms 28,536 ms 6,898 ms 12,566 ms

输入 10 10 0 10

输出 1,971 3,022 1,444 1,682

缓存创建

9,489

5m: 0 / 1h: 9,489

8,826

5m: 0 / 1h: 8,826

5m: 0 / 1h: 0

8,729

5m: 0 / 1h: 8,729

缓存读取 0 0 0 0

输出速度 154.4 tok/s 112.0 tok/s burst 159.1 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 01:09:02 ~ 01:09:17 (14.4s) 01:03:20 ~ 01:03:40 (15.3s) 01:07:32 ~ 01:07:42 (5.6s) 01:06:16 ~ 01:06:41 (20.6s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=302600464726420147744494656230377638657763905772177618113502689710828723

HTTP 头(TTFB) 1,224 ms 1,127 ms 4,444 ms 1,793 ms

首文本(SSE delta) 1,226 ms 1,128 ms 4,445 ms 1,794 ms

总耗时 13,480 ms 14,548 ms 4,574 ms 19,418 ms

输入 193 193 0 193

输出 1,980 2,587 885 3,518

缓存创建

5m: 0 / 1h: 0

缓存读取 0 0 0 0

输出速度 161.6 tok/s 192.8 tok/s burst 199.6 tok/s

响应

```json
{"n":72,"c":[8,13,3]}
```

```json
{"n":72,"c":[8,13,3]}
```

{"n":72,"c":[8,16,3]}

```json
{"n": 72, "c": [8, 13, 3]}
```

总计

总输入(含缓存) 143,382 142,867 0 142,707

总输出 5,455 6,441 2,613 5,944

总缓存创建

43,135

5m: 0 / 1h: 43,135

42,542

5m: 0 / 1h: 42,542

5m: 0 / 1h: 0

42,405

5m: 0 / 1h: 42,405

总缓存读取 100,004 100,082 0 100,059

输出速度 154.3 tok/s 133.1 tok/s burst 175.0 tok/s

总执行时间 51.8s 1m 5.8s 35.3s 55.6s

总墙钟时间

16m 34.1s

含步骤间等待

11m 1.0s

含步骤间等待

15m 7.0s

含步骤间等待

13m 60.0s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.0914

Anthropic: Claude Haiku 4.5

$0.0956

Anthropic: Claude Haiku 4.5

>=$0.0131

Anthropic: Claude Haiku 4.5 (成本不完整)

$0.0930

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%) one-api (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 - - - -

ID 格式 msg_01Fxv9B9G4srQ2xhmfVo5CSL msg_01SQQKKYvAfUZGpVqRq2N8T7 msg_8c91b4454c9b29a3498e7312 msg_01XHkqzW1VHSSqLTk5FnWgD7

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_011CcU8heK4st74aS133f1Ku
O-Max	server: nginx x-oneapi-request-id: 202606271703215996568278268d9d6w2uLaTzi	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-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: 20260627170734104551368268d9d6ildtaEhQ	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	—
M-Max	server: nginx x-new-api-version: 0b8be5cf x-oneapi-request-id: 2026062801061868989064o9Db6cJN	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	—

协议指纹评分（机器自动）

总分 / 100 100 98 60 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 10.0 /10 ×20 0.0 /10 ×20 10.0 /10 ×20

缓存 TTL 一致 10.0 /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 10.0 /10 ×14 10.0 /10 ×14 5.0 /10 ×14 10.0 /10 ×14

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

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

系统提示纯净 10.0 /10 ×8 10.0 /10 ×8 未计入 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 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 10.0 /10 ×5 10.0 /10 ×5 9.0 /10 ×5 9.5 /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） 98 v3.25.0 60 v3.25.0 98 v3.25.0

雷达图例

共 22 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 档位思考量 tier_thinking_volume_match
22 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 7624 ms 10121 ms 4665 ms 8122 ms

输出速度 154.3 tok/s 133.1 tok/s burst 175.0 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 ⓘ

$0.0914

$0.0956

>=$0.0131 (成本不完整)

$0.0930

缓存读取占比 ⓘ 70% 70% - 70%

揭晓

通道

官方基准

服务商 Novaris ↗

通道O-Max

服务商 HongMaCC ↗

通道u-cc

服务商 SudoCode ↗

通道M-Max

返回列表