Proudly Open-Source

Pricing designed for developers

Use Scoopika forever-free. only upgrade for extra features that deserve paying for

Features out of the box (free forever)

Unlimited agent runs

Connect your own LLMs

Add tools to agents

Connect agents together

Pay only for extra features

Real-time streaming

LLM output validation

Full type-safety

Smart error recovery

Seamless web integration

Get started now
No credit card required
Pricing

Hobby - Free

Get started with:
Personalized AI agents
Unlimited text runs
Real-time streaming
Get started

Basic - $16/month

Everything in Hobby, plus:
Agents can speak and listen
Persistent chat sessions
Custom knowledge sources
Get started

Scale - $56/month

Everything in Basic, plus:
Higher volumes for storage and audio
Faster speech generation
Scale to millions globally
Get started
Created agents
The number of agents you can create
Free:
2
Basic:
10
Scale:
10
Multi-agent boxes
The number of boxes you can create
Free:
1
Basic:
4
Scale:
4
Tools per Agent
The number of tools you can add to an agent
Free:
3 API tools & Unlimited functions
Basic:
Unlimited
Scale:
Unlimited
Loads
Each time your application loads an agent data from the platform (once every ~15 minutes when using the Scoopika endpoint)
Free:
5K / month
Basic:
200K / month
Scale:
1M / month
Memory
Chat sessions with users (messages history or memory)
Free:
In-memory store (not persistent)
Basic:
Serverless managed memory (1 region)
Scale:
Serverless managed memory (replicated to 4 regions)
History store reads
Each time you read a session info or messages from the serverless store (4Kb = 1 operation)
Free:
---
Basic:
1M / month
Scale:
4M / month
History store writes
Each time you new session is created or an agent is executed with history enabled to save the chat messages (1Kb = 1 operation)
Free:
---
Basic:
500K / month
Scale:
2M / month
Speech Characters
When using the agent speech (agent audio response)
Free:
50 / month
Basic:
100K / month
Scale:
1.2M / month
Knowledge requests
If using custom knowledge, with each run we feth info relevant data from a vector database using RAG (this is cached and cache doesn't count to usage)
Free:
---
Basic:
300K / month
Scale:
1.5M / month
Image inputs
Supported only with LLMs that support vision
Free:
Unlimited
Basic:
Unlimited
Scale:
Unlimited
Audio inputs
You can input audio into your agent runs. fast is 2s average latency & slow is ~10s. If limit for fast processing is reached it falls back to slow mode till next month
Free:
Unlimited slow
Basic:
500mins fast & unlimited slow
Scale:
1000mins fast & unlimited slow