6345ywz API is an OpenAI-compatible API relay providing access to multiple AI models with competitive pricing.
https://lmspeed.net/zh/provider/api-6345ywz-cn/healthSite health check
https://lmspeed.net/zh/speedAPI test
https://lmspeed.nethttps://api.6345ywz.cn/Normal route
https://api.6345ywz.cnVerify ownership to unlock provider management features:
Rankings are based on community-submitted tests and periodic health probes. Advisory only, not official data.
1.注册赠送20元，每日签到赠送0-6$随机
2.本站极不稳定，5月13日下午8点时共有166个免费模型，实测仅54个可用
3.当前注册方式：linuxDO绑定+QQ邮箱绑定注册，希望能保证一人一号，如果还是批量薅羊毛的，我就再也不开了，我这个站这么不稳定，每天给的也少，咋专薅我一个呢
4.都免费了，还要什么自行车
5.主要通过nvidia api转发，我付费加入的模型只有极少一部分
6.偶尔会出现临时渠道，临时渠道1为我自费接入其他中转站，免费提供的大量模型，但额度有限，用完即止，临时渠道2为免费中转渠道，免费，随时暴毙，可能存在提示词注入等安全风险，请注意防范
7.翻译分组速度极快，m2.7最高速度可达578tps，但请勿用于其他地方，会导致渠道失效，如有需要，请使用5倍分组，谢谢配合
·API地址：https://api.6345ywz.cn/v1
·共享翻译API：sk-RA8RUhgIi4aaZJx7oSqdjE2B66kBdV2LQ3ZDZD1cPizxiEZc
·余额不够用？（真的会不够吗？）使用LDC积分兑换！→1LDC=2刀
Compare 925 model rows across audit recency, latest speed tests, throughput, latency, and per-token pricing.
deepseek-v4-flash-0731 | - | - | ||||||||||
gpt-5.4-high | - | - | ||||||||||
gpt-5.4-low | - | - | ||||||||||
gpt-5.4-medium | - | - | ||||||||||
gpt-5.4-mini-high | - | - | ||||||||||
gpt-5.4-mini-low | - | - | ||||||||||
gpt-5.4-mini-medium | - | - | ||||||||||
gpt-5.4-mini-xhigh | - | - | ||||||||||
gpt-5.4-xhigh | - | - | ||||||||||
gpt-image-1.5 | - | - | ||||||||||
google/diffusiongemma-26b-a4b-it0倍倍率分组 | $0.014/M | $0.041/M | ||||||||||
google/diffusiongemma-26b-a4b-it1倍倍率分组 | $0.137/M | $0.411/M | ||||||||||
google/diffusiongemma-26b-a4b-it2倍倍率分组 | $0.274/M | $0.822/M | ||||||||||
google/diffusiongemma-26b-a4b-it3倍倍率分组 | $0.411/M | $1.23/M | ||||||||||
claude-fable-5-a临时渠道2 | $0.0041/M | $0.000000137/M | ||||||||||
claude-mythos-5-a临时渠道2 | $0.0041/M | $0.000000137/M | ||||||||||
claude-opus-4-6-thinking | - | - | ||||||||||
codex-auto-review | - | - | ||||||||||
gemini-3.1-flash-lite | - | - | ||||||||||
gemini-3.1-pro-preview | - | - | ||||||||||
gpt-3o-a临时渠道2 | $0.0041/M | $0.000000137/M | ||||||||||
gpt-5.4-2026-03-05 | - | - | ||||||||||
gpt-5.6-sol | - | - | ||||||||||
gpt-5.6-terra | - | - | ||||||||||
grok-imagine-image-edit | - | - |
Showing 25 of 925 model rows
Community-shared API keys and redeem codes associated with 6345ywz API.
| API key / Status | API URL | Models | Submitted |
|---|---|---|---|
x*** Checking | https://lmspeed.net/api/auth/admin/list-users?x= | Models pending | PollyAI-Official | pollyai.net· Aug 6, 2026 |
sk-7AK***ztP0 Unavailable | https://api.6345ywz.cn/v1 | Models pending | wz y· Jul 6, 2026 |
| Time | Model | Speed | Latency |
|---|---|---|---|
| Aug 22, 04:50 PM | FAST/minimax-m2.7 | 639.26 tok/s | 1.15s |
| Aug 22, 04:49 PM | FAST/minimax-m2.7 | 681.70 tok/s | 1.36s |
| Aug 22, 04:48 PM | FAST/gpt-oss-120b | 827.48 tok/s | 0.90s |
| Aug 22, 04:47 PM | FAST/minimax-m2.7 | 702.56 tok/s | 1.29s |
| Aug 3, 12:01 PM | FAST/gpt-oss-120b | 603.47 tok/s | 0.98s |
| Aug 3, 11:55 AM | FAST/minimax-m2.7 | 483.94 tok/s | 1.33s |
| Aug 1, 08:25 AM | deepseek-v4-flash | 106.10 tok/s | 2.90s |
| Jul 8, 04:14 PM | glm-5.1 | 148.11 tok/s | 5.30s |
| Jul 6, 08:15 AM | gpt-5.5 | 45.94 tok/s | 2.60s |
| Jul 6, 08:13 AM | gpt-5.5 | 43.10 tok/s | 2.73s |
Compare 6345ywz API alternatives against 6 nearby API providers using 647 LMSpeed signals across shared model coverage, pricing, benchmark speed, uptime, and free-model availability.
| Provider | Why compare | Models | Free | Avg price | Speed | 30d uptime |
|---|---|---|---|---|---|---|
| 6345ywz API api-6345ywz-cn 6345ywz API is an OpenAI-compatible API relay providing access to multiple AI models with competitive pricing. | Current provider baseline | 466 | 0 | $0.0000686/M | 272 tok/s | 9750% |
| 钠 API naapi-cc Na API (naapi.cc) is an OpenAI-compatible LLM API gateway with competitive pricing and stable access to 100+ models from OpenAI, Anthropic, Google, and more. |
| 566 | 0 | $0.0031/M | 649 tok/s | 9960% |
| 小水管 API edge-pieixan-icu Pie-Xian (小水管) provides an OpenAI-compatible API relay with multiple endpoints including api.pie-xian.com and edge nodes. |
| 495 | 2 | $0.054/M | 260 tok/s | 9900% |
| MapleLeaf API ai-071129-xyz MapleLeaf API runs a New API-powered gateway on ai.071129.xyz for aggregated access to multiple AI models. |
| 470 | 292 | N/A | 38 tok/s | 9940% |
| 初叶🍂Furry API ai-chuyel-top A free, community-supported API service providing access to various AI models for developers and users. |
| 1,021 | 152 | $18.00/M | 171 tok/s | 2360% |
| 天絮 API tianxu-api Tianxu API provides an AI model relay service with multiple access points and stable connectivity. |
| 759 | 0 | $0.014/M | 93 tok/s | 9910% |
| 速创API suchuang Suchuang API provides access to various AI models including text generation, image creation, and video generation through a unified API. |
| 510 | 0 | $0.0020/M | 74 tok/s | 9960% |
Current server is under heavy load and keeps freezing, so registration is temporarily closed.
The host machine at the server provider had issues around 1 AM and has just recovered.
1-5x channel temporarily launches GPT with a $500 limit; it will die once used up.
To handle the growing number of users, the server now automatically restarts every day at 4:00 AM Beijing time (UTC 20:00). Each restart takes about 120 seconds. Please take note if you need to use it during this time.
Discord registration and login have been disabled; reopening time unknown.
GPT channels all died last night and probably won't return.
Let me recommend a handy translation tool:
It can quickly translate untranslated text extracted by MTool from games, as well as novels, books, subtitles, documents, etc., into Chinese. It works quite well and suits batch processing.
If you've claimed a free API Key but don't know how to use it, give it a try.
The new channel test is done. The 1x group now has FAST channel and it works. If it doesn't work, the channel probably melted down—just move the multiplier up one tier.
FAST model has been added to the temporary channel; it runs on a channel that's still being tested. The cost is very low, but it's unstable and goes down from time to time—a single channel may not even survive 7 days. Feel free to try it, but if you need stability, use the 5x channel.
The shared API is being abused. 16 requests per second is terrifying—at least pretend it's normal? Who sends the exact same input token count on every request in normal use? The API key has been updated, and restrictions are now enabled for the shared API:
RPM limit 20, $3 per 5 hours, $50 per week, shared API valid for 366 days starting from this announcement.
FAST model channel has been restored.
FAST model channel blew up; no word on when it'll recover.
linuxDC login has been fixed. The previous login issue was my fault—after linuxDC's testing, I forgot to switch the callback link back. It's back to normal now. ((
As compensation, here are 5 ·20-yuan credit redeem codes.
Registration is open. It depends on server status; if it can't handle the load, registration will be closed.
Upgrade complete.
System version upgrade, expected to take 10 minutes.
Replacement complete.
Switching servers, expected to take 2 minutes.
From 12:44 to 2:15 the server was down. The single-core small machine just isn't stable enough, so we're switching to a dual-core server.
