DeepInfra raises $107M Series B to scale the inference cloud — read the announcement

Inference LoRA adapter model
Published on 2024.12.06 by Askar Aitzhan
Inference LoRA adapter model

Understanding LoRA inference

Concepts

  • Base model: The original model that is used as a starting point.
  • LoRA adapter model: A small model that is used to adapt the base model for a specific task.
  • LoRA Rank: The rank of the matrix that is used to adapt the model.

What you need to inference with LoRA adapter model

  1. Supported base model
  2. LoRA adapter model hosted on HuggingFace
  3. HuggingFace token if the LoRA adapter model is private
  4. DeepInfra account

How to inference with LoRA adapter in DeepInfra

  1. Go to the dashboard
  2. Click on the 'New Deployment' button
  3. Click on the 'LoRA Model' tab
  4. Fill the form:
    • LoRA model name: model name used to reference the deployment
    • Hugging Face Model Name: Hugging Face model name
    • Hugging Face Token: (optional) Hugging Face token if the LoRA adapter model is private
  5. Click on the 'Upload' button

Note: The list of supported base models is listed on the same page. If you need a base model that is not listed, please contact us at feedback@deepinfra.com

Rate limits on LoRA adapter model

Rate limit will apply on combined traffic of all LoRA adapter models with the same base model. For example, if you have 2 LoRA adapter models with the same base model, and have rate limit of 200. Those 2 LoRA adapter models combined will have rate limit of 200.

Pricing on LoRA adapter model

Pricing is 50% higher than base model.

How is LoRA adapter model speed compared to base model speed?

LoRA adapter model speed is lower than base model, because there is additional compute and memory overhead to apply the LoRA adapter. From our benchmarks, the LoRA adapter model speed is about 50-60% slower than base model.

How to make LoRA adapter model faster?

You could merge the LoRA adapter with the base model to reduce the overhead. And use custom deployment, the speed will be close to the base model.

Related articles
How to deploy Databricks Dolly v2 12b, instruction tuned casual language model.How to deploy Databricks Dolly v2 12b, instruction tuned casual language model.Databricks Dolly is instruction tuned 12 billion parameter casual language model based on EleutherAI's pythia-12b. It was pretrained on The Pile, GPT-J's pretraining corpus. [databricks-dolly-15k](http...
Step 3.7 Flash is Live on DeepInfra: An Agentic, Multimodal Model Built for ProductionStep 3.7 Flash is Live on DeepInfra: An Agentic, Multimodal Model Built for ProductionStepFun's Step 3.7 Flash is now live on DeepInfra. It's a 198B-parameter sparse MoE vision-language model with just ~11B active parameters per token, a 256K context window, and three selectable reasoning levels—purpose-built for high-throughput agentic workflows that combine perception, search, and reasoning.
Best API Providers for GLM-5.1 in 2026Best API Providers for GLM-5.1 in 2026<p>GLM-5.1 is available across a growing number of API providers, and the choice between them materially affects cost, latency, and what features you can actually use. The benchmark spread is real: blended pricing runs from $0.74 to $1.70 per 1M tokens across tracked providers, output speed ranges from 33 to 175 t/s, and not every [&hellip;]</p>