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-opus-4-8 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

InfAI / R-Kiro
```
评测失败
```

步骤 / 指标

官方基准

LinkAPI

O-Max

FoxCode

R-ultra

0-0

O-Max

基本信息

服务名 Claude Code Claude Code Claude Code Claude Code

模型 claude-opus-4-8 claude-opus-4-8 claude-opus-4-8 claude-opus-4-8

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 00:27:57 ~ 00:28:02 (2.8s) 00:27:51 ~ 00:27:56 (3.2s) 00:27:55 ~ 00:28:00 (3.2s) 00:27:53 ~ 00:28:03 (5.0s)

提示词

查看本步 Prompt

ping. Reply with the single word: pong

HTTP 头(TTFB) 1,509 ms 1,501 ms 1,963 ms 3,233 ms

首文本(SSE delta) 1,510 ms 1,503 ms 2,029 ms 3,234 ms

总耗时 1,696 ms 1,957 ms 2,141 ms 3,306 ms

输入 1,454 1,454 89 1,454

输出 4 4 2 4

缓存创建

30,222

5m: 0 / 1h: 30,222

30,222

5m: 30,222 / 1h: 0

23,472

5m: 23,472 / 1h: 0

30,854

5m: 0 / 1h: 30,854

缓存读取 0 0 0 0

响应

pong

pong

pong

pong

步骤 2: identity text

时间 00:29:40 ~ 00:29:45 (3.5s) 00:29:16 ~ 00:29:21 (3.8s) 00:30:32 ~ 00:30:37 (3.8s) 00:31:55 ~ 00:32:00 (4.2s)

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) 2,081 ms 1,915 ms 1,910 ms 2,494 ms

首文本(SSE delta) 2,083 ms 1,916 ms 1,911 ms 2,495 ms

总耗时 2,494 ms 2,645 ms 2,935 ms 2,952 ms

输入 2 2 15 2

输出 28 28 64 28

缓存创建

1,643

5m: 0 / 1h: 1,643

1,643

5m: 1,643 / 1h: 0

272

5m: 272 / 1h: 0

3,732

5m: 0 / 1h: 3,732

缓存读取 30,222 30,222 23,472 28,764

输出速度 - - 62.5 tok/s -

响应

vendor: anthropic
brand: claude
model: claude-opus-4-8

vendor: anthropic
brand: claude
model: claude-opus-4-8

vendor: anthropic
brand: claude
model: claude-opus-4-8

vendor: anthropic
brand: claude
model: claude-opus-4-8

步骤 3: cutoff text

时间 00:32:45 ~ 00:32:50 (2.9s) 00:31:14 ~ 00:31:19 (4.8s) 00:33:55 ~ 00:34:00 (4.4s) 00:35:08 ~ 00:35:18 (6.6s)

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) 1,538 ms 3,406 ms 2,717 ms 4,576 ms

首文本(SSE delta) 1,540 ms 3,407 ms 2,718 ms 4,577 ms

总耗时 1,841 ms 3,769 ms 3,549 ms 4,840 ms

输入 2 2 62 2

输出 6 6 36 6

缓存创建

5m: 0 / 1h: 64

5m: 62 / 1h: 0

24,090

5m: 24,090 / 1h: 0

3,795

5m: 0 / 1h: 3,795

缓存读取 31,865 31,865 0 28,764

输出速度 - - 43.3 tok/s -

响应

2026-01

2025-01

2026-01

2026-01

步骤 4: identity_free text

时间 00:36:26 ~ 00:36:31 (4.0s) 00:34:04 ~ 00:34:09 (3.7s) 00:37:13 ~ 00:37:18 (4.4s) 00:38:02 ~ 00:38:07 (4.7s)

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) 2,082 ms 1,581 ms 2,052 ms 2,679 ms

首文本(SSE delta) 2,084 ms 1,582 ms 2,053 ms 2,680 ms

总耗时 2,916 ms 2,599 ms 3,563 ms 3,253 ms

输入 2 2 16 2

输出 46 59 64 47

缓存创建

5m: 0 / 1h: 37

5m: 31 / 1h: 0

24,514

5m: 24,514 / 1h: 0

3,826

5m: 0 / 1h: 3,826

缓存读取 31,929 31,927 0 28,764

输出速度 55.3 tok/s 58.0 tok/s 42.4 tok/s 82.0 tok/s

响应

我是 Claude，由 Anthropic 开发，属于 Claude 模型系列，当前型号为 Opus 4.8。

我是 Claude,一个由 Anthropic 开发的人工智能助手。具体型号是 Claude Opus 4.8,可以帮助你完成各种任务。

我叫 Claude，是 Anthropic 开发的 AI 助手，当前使用的模型是 claude-opus-4-8。

我是 Claude，一个由 Anthropic 开发的人工智能助手，具体为 Claude Opus 4.8 模型。

步骤 5: knowledge_recall text

时间 00:40:13 ~ 00:40:28 (10.2s) 00:37:50 ~ 00:38:00 (5.8s) 00:40:47 ~ 00:41:17 (29.0s) 00:39:31 ~ 00:40:46 (70.1s)

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) 1,436 ms 1,253 ms 1,301 ms 11,121 ms

首文本(SSE delta) 8,776 ms 2,221 ms 1,302 ms 65,466 ms

总耗时 9,251 ms 4,718 ms 28,212 ms 68,671 ms

输入 2 2 63 2

输出 701 386 2,414 327

缓存创建

4,766

5m: 0 / 1h: 4,766

4,771

5m: 4,771 / 1h: 0

4,675

5m: 4,675 / 1h: 0

5,342

5m: 0 / 1h: 5,342

缓存读取 0 0 0 0

输出速度 - burst 89.7 tok/s 102.0 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 00:41:55 ~ 00:42:00 (3.0s) 00:39:29 ~ 00:39:34 (3.0s) 00:44:07 ~ 00:44:27 (18.2s) 00:43:10 ~ 00:43:15 (4.4s)

提示词

查看本步 Prompt

Count digits exactly in s. Return {"n":len(s),"c":[count0,count7,count9]}. s=304788750940477682333977519222546958673540798218214374045603465703559841

HTTP 头(TTFB) 1,706 ms 1,001 ms 1,689 ms 2,126 ms

首文本(SSE delta) 1,708 ms 1,002 ms 1,690 ms 2,839 ms

总耗时 2,086 ms 1,928 ms 17,389 ms 2,840 ms

输入 211 211 22 793

输出 19 19 1,921 19

缓存创建

5m: 0 / 1h: 0

187

5m: 187 / 1h: 0

5m: 0 / 1h: 0

缓存读取 0 0 0 0

输出速度 - - 122.4 tok/s -

响应

{"n":72,"c":[7,12,5]}

{"n":72,"c":[5,12,4]}

{"n":72,"c":[7,10,6]}

{"n":71,"c":[6,9,4]}

总计

总输入(含缓存) 132,421 132,416 100,949 136,096

总输出 804 502 4,501 431

总缓存创建

36,732

5m: 0 / 1h: 36,732

36,729

5m: 36,729 / 1h: 0

77,210

5m: 77,210 / 1h: 0

47,549

5m: 0 / 1h: 47,549

总缓存读取 94,016 94,014 23,472 86,292

输出速度 55.3 tok/s 58.0 tok/s 97.9 tok/s 99.0 tok/s

总执行时间 26.3s 24.1s 1m 3.0s 1m 35.0s

总墙钟时间

14m 2.9s

含步骤间等待

11m 43.0s

含步骤间等待

16m 32.1s

含步骤间等待

15m 21.2s

含步骤间等待

按官方价目重估

同等内容若直发 Anthropic 的估算 ⓘ

$0.3050

Anthropic: Claude Opus 4.8

$0.2114

Anthropic: Claude Opus 4.8

$0.4272

Anthropic: Claude Opus 4.8

$0.3624

Anthropic: Claude Opus 4.8

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%) 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_01Qmm4ydj3PYSLxi23eqnWoA msg_01FdoTt7u4og2dVAd4ZN4uwb msg_01JDZjwtQgrYy0QZwFKf8oyn msg_019S8csumrDVyJ7oFhUCZzZf

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: 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_011CcVzSdsBsW2Qyr1LVQpcq
O-Max	server: nginx set-cookie: server_name_session=7d88fca8b1426ca76c58d5aeba6fb224; Max-Age=86400; httponly; path=/ x-new-api-version: v1.0.0-rc.15 x-oneapi-request-id: 202606281639309265736548268d9d6SSXiBkka	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	—
R-ultra	cf-ray: a12e278a0d5b7963-NRT server: cloudflare x-new-api-version: v0.0.0 x-oneapi-request-id: 202606281644088275719238268d9d6autkrx7w	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	a12e278a0d5b7963-NRT
O-Max	via: 1.1 Caddy server: uvicorn, nginx/1.24.0 (Ubuntu)	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: false usage_has_cache_fields: true	hdr:viabody:id:msg_prefixbody:usage.service_tierbody:usage.inference_geobody:usage.cache_fieldsbody:extra_fieldsupstream:anthropic-direct-from-body	e3bd4567dd533cdf28e0e7335586292e

协议指纹评分（机器自动）

总分 / 100 97 81 59 92

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 0.0 /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 10.0 /10 ×14 10.0 /10 ×14

sliding 5m 10.0 /10 ×13 10.0 /10 ×13 0.0 /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 0.0 /10 ×8 5.0 /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 0.0 /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 8.2 /10 ×5

流式投递 0.0 /10 ×5 10.0 /10 ×5 6.6 /10 ×5 10.0 /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

综合结论

与基线相似度基准（参考） 81 v3.25.0 59 v3.25.0 92 v3.25.0

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 系统提示纯净 system_prompt_clean
9 Service Tier service_tier_present
10 缓存命中比 cache_hit_ratio_match
11 缓存 TTL 一致 cache_ttl_consistency
12 推理地区 inference_geo_present
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 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 3381 ms 2936 ms 9632 ms 14310 ms

输出速度 55.3 tok/s 58.0 tok/s 97.9 tok/s 99.0 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 ⓘ

$0.3050

$0.2114

$0.4272

$0.3624

缓存读取占比 ⓘ 71% 71% 23% 63%

揭晓

通道

官方基准

服务商 LinkAPI ↗

通道O-Max

服务商 FoxCode ↗

通道R-ultra

服务商 0-0 ↗

通道O-Max

返回列表