GPT Load (Shiho) is an OpenAI-compatible API load balancing service hosted at gpt-load.shiho.top, distributing requests across multiple AI model providers for improved reliability.
GPT Load (Shiho) offers 22 LLM API models.
Speed benchmark average: 1164 tok/s.
gpt-load.shiho.topRankings are based on community-submitted tests and periodic health probes. Advisory only, not official data.
| Model | Speed | Latency | Tests |
|---|---|---|---|
481.90 tok/s | 0.43s | 5 | |
1629.33 tok/s | 0.36s | 25 | |
| Time | Model | Speed | Latency |
|---|---|---|---|
| Jun 3, 06:58 PM | openai/gpt-oss-120b | 481.90 tok/s | 0.43s |
| Dec 25, 02:06 PM | llama3.1-8b | 2142.09 tok/s | 0.19s |
| Dec 25, 02:02 PM | llama-3.3-70b | 1374.74 tok/s | 0.25s |
| Sep 21, 06:22 PM | llama3.1-8b | 1834.10 tok/s | 0.35s |
| Sep 21, 06:21 PM | llama-4-maverick-17b-128e-instruct | 825.70 tok/s | 0.41s |
| Sep 21, 06:19 PM | llama-3.3-70b | 890.30 tok/s | 0.51s |
| Sep 21, 06:18 PM | qwen-3-235b-a22b-thinking-2507 | 579.82 tok/s | 0.44s |
| Sep 21, 06:17 PM | llama3.1-8b | 2117.91 tok/s | 0.34s |
| Jun 8, 11:00 PM | llama3.1-8b | 806.21 tok/s | 0.47s |
| Jun 8, 10:59 PM | llama-3.3-70b | 794.97 tok/s | 0.52s |
984.90 tok/s |
0.45s |
| 20 |
825.70 tok/s | 0.41s | 5 |
qwen-3-235b-a22b-thinking-2507 | 579.82 tok/s | 0.44s | 5 |
newapi.ixio.cc
IXIOCCAPI is a unified API gateway for large language models, providing standardized endpoints for accessing multiple AI model providers.
chat-api4.087654.xyz
天絮 API provides an AI model relay service with multiple access points and stable connectivity.
newapi.exynos.top:8443
Synapse is an OpenAI-compatible API relay service providing access to multiple AI models with unified endpoints.
napi.seaya.link
Seamee API provides an AI model relay for accessing multiple LLMs through OpenAI-compatible endpoints.
apifree.rensumo.top
A non-profit AI infrastructure offering free access to integrated large language models with privacy-focused, no-logging policies.
api.hotaruapi.top
HotaruAPI provides API access to AI models for developers, including a model marketplace and self-service options.
