Get started with Airtrain for free

LLM Playground

Query multiple models at once. Compare across 23 open-source and proprietary models.

LLM playground
Query multiple models at once
23 models supported
Compare quality, cost, throughput
Inference metrics
Persisted sessions
Support on Slack
Evaluation

Explore, compare, and evaluate open-source and proprietary LLMs on your own tasks and datasets.

Batch LLM evaluation
Compare open-source and proprietary LLMs on your own dataset
LLM-assisted scoring
JSON schema validation
Unsupervised metrics
Support on Slack
Free

Free for datasets of up to 10,000
examples. Need more? Get in touch.

Get started
Fine-tuning

Customize pre-trained LLMs for your own applications and save up to 90% compared to proprietary solutions.

Fine-tune open-source LMs
Upload your own dataset
No-code job parametrization
Cost- and speed-optimized
infrastructure
Export final model for serving
Support on Slack
Per token

Predictable upfront price.

Get started
Enterprise

Build and execute training and evaluation
workloads on your own cloud infrastructure.

Deployed in your private and secure cloud
environment.infrastructure.

Open-source compute infrastructure
On-premise deployment
Template pipelines for training, fine-
tuning, evaluation
Python SDK to build custom pipelines
Export final model for serving
Dedicated support with Slack Connect
Get in touch

Free for datasets of up to 10,000
examples.Need more? Get in touch.

Book a call
Effortless Comparison

Effortlessly compare open-source models

Effortlessly evaluate open-source models on your own data, define your own criteria, and choose the best fit for your use case.

Llama 2
Llama 2
7B
13B
70B
Mistral 7B
Mistral
7B
8x7B
Medium
Custom Made
Phi-2
2.7B
FLAN-T5
Gemini
Pro
Ultra
Open AI
Open AI
GPT-3.5
GPT-4
Falcon
Claude 3
Opus
Sonnet
Haiku

Customize LLMs for your unique tasks

Each use case is unique. Airtrain lets you score, evaluate, and tune large language models on your dataset for your task.

Get started