Overview
AI-driven inference for LLMs and multimodal applications.
Siliconflow is an AI infrastructure platform for inference across large-language models (LLMs) and multimodal applications.It supports serverless, reserved, and private-cloud deployment.Features include high-speed inference for image and video processing, support for various LLMs, and fine-tuning capabilities.
The platform offers low latency, high throughput, predictable costs, and built-in monitoring with elastic compute.
Use Cases
- Leverage Siliconflow to deploy large-language models for real-time customer support chatbots that provide accurate and context-aware responses without latency issues.
- Utilize Siliconflow's image and video processing capabilities to create an AI-driven content moderation system that automatically identifies and flags inappropriate content across multimedia platforms.
- Implement Siliconflow's fine-tuning features to customize AI models for specific industry needs, such as enhancing predictive analytics in finance or personalized recommendations in e-commerce, all while ensuring cost predictability and efficient resource management.
Pricing
Free plan available. Paid plans unlock more usage.
- Flux 1.1 [pro]: $0.04
- Flux.1 kontext [pro]: $0.04
- Flux 1.1 [pro] ultra: $0.06
- Flux.1 kontext [max]: $0.08
- Qwen3-embedding-0.6b: $0.01/$0
- Qwen3-reranker-0.6b: $0.01/$0
- Flux.1-dev: $0.014
- Flux.1-schnell: $0.0014
- Flux.1-kontext-dev: $0.015
- Fish-speech-1.5: $15
- Qwen3-embedding-4b: $0.02/$0
- Qwen3-reranker-4b: $0.02/$0
- Wan2.1-i2v-14b-720p (turbo): $0.21
- Wan2.1-t2v-14b (turbo): $0.21
- Wan2.1-i2v-14b-720p: $0.29
- Wan2.1-t2v-14b: $0.29
- Qwen3-embedding-8b: $0.04/$0
- Qwen3-reranker-8b: $0.04/$0
- Glm-4.5: $0.5/$2
- Deepseek-r1-distill-qwen-14b: $0.1/$0.1
- Qwen2.5-14b-instruct: $0.1/$0.1
- Qwen3-30b-a3b: $0.1/$0.4
- Qwen3-30b-a3b-instruct-2507: $0.1/$0.4
- Qwen3-30b-a3b-thinking-2507: $0.1/$0.4
- Qwen3-coder-30b-a3b-instruct: $0.1/$0.4
- Funaudiollm/cosyvoice2-0.5b: $7.15
- Deepseek-r1-distill-qwen-7b: $0.05/$0.05
- Qwen2.5-7b-instruct: $0.05/$0.05
- Qwen2.5-vl-7b-instruct: $0.05/$0.05
- Meta-llama-3.1-8b-instruct: $0.06/$0.06
- Qwen3-8b: $0.06/$0.06
- Qwen3-14b: $0.07/$0.28
- Qwen3-32b: $0.14/$0.57
- Hunyuan-a13b-instruct: $0.14/$0.57
- Glm-z1-32b-0414: $0.14/$0.57
- Glm-4.5-air: $0.14/$0.86
- Deepseek-vl2: $0.15/$0.15
- Qwq-32b: $0.15/$0.58
- Deepseek-r1-distill-qwen-32b: $0.18/$0.18
- Qwen2.5-32b-instruct: $0.18/$0.18
- Qwen2.5-coder-32b-instruct: $0.18/$0.18
- Qwen2.5-vl-32b-instruct: $0.27/$0.27
- Glm-4-32b-0414: $0.27/$0.27
- Ernie-4.5-300b-a47b: $0.29/$1.15
- Deepseek-v3: $0.29/$1.15
- Glm-4.1v-9b-thinking: $0.035/$0.14
- Qwen3-235b-a22b: $0.35/$1.42
- Qwen3-235b-a22b-2507: $0.35/$1.42
- Qwen3-235b-a22b-thinking-2507: $0.35/$1.42
- Step3: $0.57/$1.42
- Deepseek-r1: $0.58/$2.29
- Minimax-m1-80k: $0.58/$2.29
- Kimi-k2-instruct: $0.58/$2.29
- Qwen2.5-72b-instruct: $0.59/$0.59
- Qwen2.5-72b-instruct-128k: $0.59/$0.59
- Qwen2.5-vl-72b-instruct: $0.59/$0.59
- Qwen3-coder-480b-a35b: $1.14/$2.28
- Glm-4-9b-0414: $0.086/$0.086
- Glm-z1-9b-0414: $0.086/$0.086
Who Is It For
- Software developers
- Cloud architects
- Data analysts
- System administrators
- Infrastructure engineers
