提供免费的、兼容OpenAI的API接口，支持多个开源大语言模型，无需注册即可使用。
该服务提供免费的、兼容OpenAI接口的API，支持多个开源大语言模型。主要模型包括DeepSeek-R1-0528、DeepSeek-V3-0324、Qwen3-30B-A3B、Qwen3-14B、Minimax-M2和GLM-4.1V-Thinking。API支持前端跨域调用，适合AI编程，可直接使用现有的OpenAI SDK。支持流式响应。无硬性使用限制，但会监控使用情况以防止滥用。典型用例包括翻译、代码辅助和多模态任务如图像理解。服务仅限非商业用途，不保证可用性或内容准确性。
| Model | Speed | Latency | Tests |
|---|---|---|---|
| deepseek/deepseek-v3 | 172.09 t/s | 2.64s | 90 |
| deepseek/deepseek-v3-0324 | 166.30 t/s | 3.98s | 400 |
| zhipu/glm-4.1v-thinking-flash | 107.35 t/s | 7.54s | 35 |
| zhipu/glm-4.1v-thinking-flash | 107.35 t/s | 7.54s | 35 |
| openai/gpt-oss-20b | 94.04 t/s | 2.02s | 120 |
| qwen/qwen2.5-7b | 72.53 t/s | 1.95s | 50 |
| qwen/qwen2.5-7b | 72.53 t/s | 1.95s | 50 |
| zhipu/glm-4-9b | 65.28 t/s | 0.87s | 10 |
| qwen/qwen3-30b-a3b | 64.27 t/s | 7.53s | 95 |
| zhipu/glm-4v-flash | 56.90 t/s | 0.64s | 45 |
| qwen/qwen2.5-72b | 48.71 t/s | 1.46s | 65 |
| qwen/qwen2.5-72b | 48.71 t/s | 1.46s | 65 |
| google/gemma-3-27b | 39.48 t/s | 2.25s | 145 |
| google/gemini-2.0-flash-exp | 38.94 t/s | 0.79s | 310 |
| meituan/longcat-flash-chat | 36.04 t/s | 2.48s | 25 |
| zhipu/glm-4-flash | 35.05 t/s | 0.84s | 4790 |
| deepseek/deepseek-r1-70b | 29.42 t/s | 2.43s | 65 |
| zhipu/glm-4.5-flash | 28.96 t/s | 13.06s | 60 |
| zhipu/glm-4.5-flash | 28.96 t/s | 13.06s | 60 |
| deepseek/deepseek-r1 | 27.87 t/s | 4.85s | 145 |
| qwen/qwen3-14b | 27.16 t/s | 6.30s | 15 |
| qwen/qwen3-235b-a22b | 25.81 t/s | 12.66s | 10 |
| meta/llama-4-scout | 21.43 t/s | 0.81s | 35 |
| qwen/qwen3-32b | 18.38 t/s | 5.94s | 35 |
| tencent/hunyuan-a13b | 16.73 t/s | 4.38s | 40 |
| deepseek/deepseek-r1-32b | 16.25 t/s | 1.34s | 100 |
| openrouter/optimus-alpha | 15.51 t/s | 1.72s | 25 |
| qwen/qwen2.5-vl-32b | 13.61 t/s | 0.78s | 30 |
| qwen/qwen2.5-vl-32b | 13.61 t/s | 0.78s | 30 |
| qwen/qwen2.5-vl-32b | 13.61 t/s | 0.78s | 30 |
| minimax/minimax-m2 | 11.86 t/s | 2.19s | 90 |
| zhipu/glm-z1-32b | 10.97 t/s | 6.40s | 40 |
| qwen/qwen2.5-vl-72b | 9.93 t/s | 1.16s | 30 |
| qwen/qwen2.5-vl-72b | 9.93 t/s | 1.16s | 30 |
| qwen/qwen2.5-vl-72b | 9.93 t/s | 1.16s | 30 |
| deepseek/deepseek-r1-0528 | 9.27 t/s | 2.79s | 230 |
| qwen/qwen3-coder | 8.78 t/s | 0.78s | 120 |
| qwen/qwen3-coder | 8.78 t/s | 0.78s | 120 |
| moonshotai/kimi-k2 | 7.42 t/s | 0.56s | 185 |
| zhipu/glm-4-32b | 6.89 t/s | 1.10s | 25 |
| google/gemini-2.5-pro-exp | 3.33 t/s | 1.01s | 245 |
| deepseek/deepseek-chat-v3.1 | 0.00 t/s | 0.00s | 60 |
| meta/llama-4-maverick | 0.00 t/s | 0.68s | 50 |
| openrouter/quasar-alpha | 0.00 t/s | 0.35s | 30 |
| Time | Model | Speed | Latency |
|---|---|---|---|
| Dec 11, 07:37 PM | deepseek/deepseek-v3-0324 | 0.00 t/s | 0.00s |
| Dec 11, 07:36 PM | deepseek/deepseek-chat-v3.1 | 0.00 t/s | 0.00s |
| Dec 11, 07:35 PM | deepseek/deepseek-r1-32b | 0.00 t/s | 0.00s |
| Dec 11, 07:35 PM | deepseek/deepseek-v3-0324 | 0.00 t/s | 0.00s |
| Dec 11, 06:11 PM | meituan/longcat-flash-chat | 0.00 t/s | 0.00s |
| Dec 11, 01:51 AM | zhipu/glm-4-flash | 26.79 t/s | 1.98s |
| Dec 10, 01:37 AM | Unknown | - | -s |
| Dec 9, 09:12 PM | Unknown | - | -s |
| Dec 9, 12:55 PM | Unknown | - | -s |
| Dec 9, 09:19 AM | Unknown | - | -s |