Create AI apps with the
Model
Input / 1K tokens
Ouput / 1K tokens
GPT4
4 credits
7 credits
GPT3.5-Turbo
0.1 credits
0.2 credits
Claude Instant
0.2 credit
0.5 credits
Llama 2
0 credits
5 credits
Gemini
TBD
TBD
Claude 2
0.8 credits
2.4 credits
Model
Speed / 1K tokens
Overall (MMLU)
GPT4
76 seconds
87.3%
Claude Instant
9 seconds
73.4%
Gemini Pro
18 seconds
79.1%
Claude 2
31 seconds
78.5%
Llama 2
132 seconds
68.0%
GPT3.5-Turbo
26 seconds
70.0%
Model
Overall (MMLU)
Coding (Python)
GPT3.5-Turbo
70.0%
48.1%
Gemini Ultra
90.0%
74.4%
GPT4
87.3%
67.0%
Claude 2
78.5%
70.0%
Llama 2
68.0%
29.9%
Gemini Pro
79.1%
67.7%
best performance
One-click A/B test inputs, models, and prompts
Train your AI with your own data
Know when to use each LLM for your use case
One-Click Deploy
Connect to your app via API and deploy in one-click. Create unlimited drafts and re-deploy continuously.
API
Version Control
Advanced Logic
Have LLMs run your app's logic. Describe what you'd like to happen and watch your app run.
Functions
LLM Decision-Making
Prompt Engineering
A/B test prompts, models, and inputs in one click. Use pre-written snippets to enhance your prompts.
A/B Testing
Prompt Snippet
Inputs
Connect Your Data
Upload your data and get custom LLM outputs for your specific use case.
Any File
Vector Databases
App 1
Release
App 2
Release
App 3
Release
LLM Decision-Maker
Function
Function
A/B Test
Same Prompt
Llama Output
Same Prompt
Claude Output
Your Data
Public Data
APIs
Your App
Use every top LLM
Always Use the Best Model for Your App
We'll help you know when to use each model so you're saving money and optimizing performance.
