About DeepInfra
DeepInfra is an inference service that exposes open-source machine learning models through simple APIs. Developers pay for what they use instead of renting servers, and the service handles hosting, scaling and uptime for the models.
The catalog is organized by task, including text generation, embeddings, rerankers, automatic speech recognition, text to image, music, speech and video, world models and zero-shot image classification. It lists model families such as DeepSeek, Flux, Kimi, Llama, Mistral, Nemotron and Qwen. Beyond serverless models, the vendor offers on-demand GPU instances, a DeepStart program and DeepCluster for larger deployments, plus a chat interface and documentation.
DeepInfra is a proprietary commercial service. Language models are priced per million tokens and rented GPU instances by the hour, with current rates published on its pricing page, and sales contact is available for larger commitments.
Key features
- Pay-per-use APIs for open models
- Text, image, speech, music and video models
- Embeddings and reranker models
- On-demand GPU instance rental
- DeepStart and DeepCluster programs
- Chat interface for trying models
Good fit for
- Calling open LLMs without managing GPUs
- Generating embeddings for search
- Renting GPU instances by the hour
DeepInfra: questions and answers
- What is DeepInfra used for?
- DeepInfra is a pay-per-use API for running open-source AI models, covering text generation, embeddings, speech, image and video models, with rentable GPUs. It is a good fit for calling open LLMs without managing GPUs, generating embeddings for search and renting GPU instances by the hour.
- How much does DeepInfra cost?
- DeepInfra is a paid product with no free plan. Pay per million tokens by model, for example $0.02 input and $0.04 output for Llama 3.1 8B Turbo, $0.10 and $0.32 for Llama 3.3 70B Turbo, and $3.00 and $15.00 for claude-sonnet-5.
- Is DeepInfra open source?
- No. DeepInfra is proprietary (closed-source) software and can't be self-hosted. Open-source alternatives to DeepInfra include SGLang, Ollama and vLLM.
- What are some alternatives to DeepInfra?
- DeepInfra competes with Together AI, Fireworks AI and Replicate. For open-source options, see Enlisted's ranked list of open-source DeepInfra alternatives.
Open-source alternatives to DeepInfra
See all
SGLang
AI Infrastructure
SGLang is a high-performance serving framework for large language models and multimodal mo
Apache-2.0vs Amazon Bedrock★ 39k
Ollama
AI Infrastructure
Get up and running with Kimi, GLM, MiniMax, DeepSeek, gpt-oss, Qwen, Gemma and other model
MITvs ChatGPT★ 187k
vLLM
AI Infrastructure
A high-throughput and memory-efficient inference and serving engine for LLMs
Apache-2.0vs Amazon Bedrock★ 97k
LiteLLM
AI Infrastructure
The fastest, litest AI Gateway. Rust core with Python SDK. Call 100+ LLM APIs in OpenAI (o
OSSvs Amazon Bedrock★ 63k
LocalAI
AI Infrastructure
LocalAI is the open-source AI engine. Run any model - LLMs, vision, voice, image, video -
MITvs ChatGPT★ 50k
Helicone
AI Infrastructure
🧊 Open source LLM observability platform. One line of code to monitor, evaluate, and expe
Apache-2.0vs OpenRouter★ 6.3k
SaaS alternatives to DeepInfra
See all
Together AI
AI Infrastructure
Cloud platform for running, fine-tuning and training open and custom AI models
SaaS
Fireworks AI
AI Infrastructure
Inference platform for running and fine-tuning generative AI models at low latency
SaaS
Replicate
AI Infrastructure
API platform for running open-source AI models in the cloud
SaaS
Baseten
AI Infrastructure
Inference platform for deploying and scaling AI models in production
SaaS
Groq
AI Infrastructure
AI inference cloud and API built on custom LPU hardware for fast model serving
SaaS
OpenRouter
AI Infrastructure
Unified API gateway for accessing many large language models from one endpoint
SaaS

Anyscale Endpoints
Modal
RunPod
Banana
Forefront
Mistral AI
Cerebras
Portkey