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

2026/06/24 运行日期

claude-haiku-4-5-20251001 官方基准

1 任务失败执行不完整

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

duckcoding / cc专用
```
评测失败
```

步骤 / 指标

Aiide

O-Max

官方基准

Novaris

O-Max

AiTokensFlux

O-Pro

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 02:28:19 ~ 02:28:29 (7.2s) 02:28:23 ~ 02:28:33 (4.9s) 02:28:17 ~ 02:28:22 (4.4s) 02:28:15 ~ 02:28:25 (6.4s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 4,088 ms 2,918 ms 2,052 ms 3,430 ms

首文本(SSE delta) 4,092 ms 2,919 ms 2,053 ms 3,432 ms

总耗时 5,853 ms 3,300 ms 3,188 ms 4,016 ms

输入 10 10 10 10

输出 57 52 57 49

缓存创建

1,635

5m: 0 / 1h: 1,635

33,122

5m: 0 / 1h: 33,122

33,122

5m: 0 / 1h: 33,122

33,261

5m: 0 / 1h: 33,261

缓存读取 31,626 0 0 0

输出速度 32.4 tok/s 136.5 tok/s 50.2 tok/s 83.9 tok/s

响应

pong

pong

pong

pong

步骤 2: identity text

时间 02:31:24 ~ 02:31:34 (8.6s) 02:30:59 ~ 02:31:09 (5.5s) 02:32:13 ~ 02:32:23 (7.4s) 02:32:10 ~ 02:32:20 (8.6s)

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。请按下列 3 行格式精确填空作答（label 不变，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) 3,455 ms 2,368 ms 2,227 ms 2,832 ms

首文本(SSE delta) 3,458 ms 2,380 ms 2,228 ms 2,880 ms

总耗时 7,526 ms 4,270 ms 5,847 ms 7,334 ms

输入 10 10 10 10

输出 140 229 194 223

缓存创建

241

5m: 0 / 1h: 241

236

5m: 0 / 1h: 236

241

5m: 0 / 1h: 241

233

5m: 0 / 1h: 233

缓存读取 33,261 33,122 33,122 33,261

输出速度 34.4 tok/s 121.2 tok/s 53.6 tok/s 50.1 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 02:33:45 ~ 02:33:55 (5.9s) 02:33:37 ~ 02:33:43 (3.9s) 02:34:18 ~ 02:34:23 (4.4s) 02:34:28 ~ 02:34:33 (4.9s)

提示词

查看本步 Prompt

请直接回答你的知识截止日期，格式 YYYY-MM，不要附加任何说明。

HTTP 头(TTFB) 3,411 ms 2,363 ms 2,371 ms 3,420 ms

首文本(SSE delta) 3,412 ms 2,365 ms 2,373 ms 3,422 ms

总耗时 4,688 ms 2,924 ms 3,639 ms 4,028 ms

输入 10 10 10 10

输出 99 77 65 65

缓存创建

182

5m: 0 / 1h: 182

269

5m: 0 / 1h: 269

217

5m: 0 / 1h: 217

246

5m: 0 / 1h: 246

缓存读取 33,502 33,358 33,363 33,494

输出速度 77.6 tok/s 137.7 tok/s 51.3 tok/s 107.3 tok/s

响应

2025-02

2025-02

2025-02

2025-02

步骤 4: identity_free text

时间 02:35:06 ~ 02:35:16 (9.4s) 02:36:08 ~ 02:36:18 (6.7s) 02:35:53 ~ 02:36:03 (7.8s) 02:36:53 ~ 02:36:58 (4.3s)

提示词

查看本步 Prompt

中文回答：你叫什么、属于哪个产品、谁开发？50 字内。

HTTP 头(TTFB) 3,980 ms 3,591 ms 2,806 ms 2,219 ms

首文本(SSE delta) 3,988 ms 3,592 ms 2,821 ms 2,220 ms

总耗时 8,327 ms 5,612 ms 6,994 ms 3,358 ms

输入 10 10 10 10

输出 323 305 348 153

缓存创建

139

5m: 0 / 1h: 139

111

5m: 0 / 1h: 111

107

5m: 0 / 1h: 107

105

5m: 0 / 1h: 105

缓存读取 33,684 33,627 33,580 33,740

输出速度 74.4 tok/s 151.0 tok/s 83.4 tok/s 134.4 tok/s

响应

我是Claude Haiku 4.5，属于Claude产品线，由Anthropic公司开发的AI助手。

我是Claude，由Anthropic公司开发的AI助手。我可以帮助用户完成各种任务，包括编程、写作、分析等。

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、写作和回答问题。

我是Claude Haiku 4.5，由Anthropic公司开发的AI助手。属于Claude产品系列，是其轻量级高效版本。

步骤 5: knowledge_recall text

时间 02:38:35 ~ 02:38:50 (13.8s) 02:37:20 ~ 02:37:50 (27.7s) 02:38:48 ~ 02:39:58 (66.8s) 02:39:52 ~ 02:40:12 (18.0s)

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) 3,028 ms 2,330 ms 3,987 ms 2,344 ms

首文本(SSE delta) 3,029 ms 2,331 ms 3,995 ms 2,346 ms

总耗时 12,726 ms 26,614 ms 65,967 ms 17,123 ms

输入 10 10 10 10

输出 1,645 3,838 4,445 2,314

缓存创建

8,625

5m: 0 / 1h: 8,625

8,661

5m: 0 / 1h: 8,661

8,653

5m: 0 / 1h: 8,653

2,756

5m: 0 / 1h: 2,756

缓存读取 0 0 0 5,721

输出速度 169.6 tok/s 158.1 tok/s 71.7 tok/s 156.6 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

```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 02:40:45 ~ 02:41:15 (28.3s) 02:40:47 ~ 02:41:02 (14.1s) 02:42:21 ~ 02:42:47 (21.1s) 02:41:13 ~ 02:42:08 (51.5s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=406873501474168074747688098147495075897100805105335503989762518295212981

HTTP 头(TTFB) 1,604 ms 898 ms 1,055 ms 3,037 ms

首文本(SSE delta) 1,605 ms 900 ms 1,056 ms 3,061 ms

总耗时 27,304 ms 12,943 ms 20,460 ms 50,710 ms

输入 193 193 193 193

输出 5,212 1,595 3,664 3,659

缓存创建

5m: 0 / 1h: 0

缓存读取 0 0 0 0

输出速度 202.8 tok/s 132.4 tok/s 188.8 tok/s 76.8 tok/s

响应

```json
{"n": 72, "c": [10, 9, 7]}
```

```json
{"n": 72, "c": [10, 9, 7]}
```

```json
{"n": 72, "c": [10, 9, 7]}
```

```json
{"n": 72, "c": [10, 9, 7]}
```

总计

总输入(含缓存) 143,138 142,749 142,648 143,060

总输出 7,476 6,096 8,773 6,463

总缓存创建

10,822

5m: 0 / 1h: 10,822

42,399

5m: 0 / 1h: 42,399

42,340

5m: 0 / 1h: 42,340

36,601

5m: 0 / 1h: 36,601

总缓存读取 132,073 100,107 100,065 106,216

输出速度 159.6 tok/s 148.0 tok/s 95.8 tok/s 93.4 tok/s

总执行时间 1m 13.4s 1m 2.7s 1m 51.9s 1m 33.7s

总墙钟时间

12m 56.0s

含步骤间等待

12m 38.9s

含步骤间等待

14m 29.4s

含步骤间等待

13m 53.0s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.0644

Anthropic: Claude Haiku 4.5

$0.0937

Anthropic: Claude Haiku 4.5

$0.1070

Anthropic: Claude Haiku 4.5

$0.0889

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%) direct (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 - - -

ID 格式 msg_01PhL95W2zR4vaFaobwjmscE msg_01Pcqf8E29pEs98ZNKEG3RK1 msg_015FdCo5Pu2GtgUNvufo2iVu msg_01WJxti8ckUUTWoMqj3RwQJk

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: a1059f82ab7f77ce-AMS server: cloudflare x-new-api-version: 0.12.11 x-oneapi-request-id: 202606231840467487230598268d9d6IQ7mNTU1	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	a1059f82ab7f77ce-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_011CcLgTkfNCERgaubr3wX8x
O-Max	server: nginx x-oneapi-request-id: 202606231842226361126728268d9d65Cwtu9PP	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	—
O-Pro	server: nginx	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	—

协议指纹评分（机器自动）

总分 / 100 98 100 98 97

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 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 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 10.0 /10 ×14 10.0 /10 ×14

sliding 5m 10.0 /10 ×13 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 10.0 /10 ×8

系统提示纯净 10.0 /10 ×8 10.0 /10 ×8 10.0 /10 ×8 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 10.0 /10 ×5 10.0 /10 ×5

Req-ID 透传 0.0 /10 ×4 10.0 /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 98 v3.24.0 基准（参考） 98 v3.24.0 97 v3.24.0

雷达图例

共 23 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 中文遵循 instruction_following_lang
22 档位思考量 tier_thinking_volume_match
23 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 11071 ms 9277 ms 17683 ms 14428 ms

输出速度 159.6 tok/s 148.0 tok/s 95.8 tok/s 93.4 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 ⓘ

$0.0644

$0.0937

$0.1070

$0.0889

缓存读取占比 ⓘ 92% 70% 70% 74%

揭晓

通道

服务商 Aiide ↗

通道O-Max

官方基准

服务商 Novaris ↗

通道O-Max

服务商 AiTokensFlux ↗

通道O-Pro

返回列表