Pay the model rate, up to your circuit breaker
Pay the model provider's rate for inference with no markup, and set a ceiling that stops AI spend at the amount that you choose.
Launch prices by currency
AI allowances per plan
Every plan includes an inference budget and a quota of policy checks, context queries, and indexed storage in each billing cycle.
Per billing cycle
Build
Pro
Business
Enterprise
Inference budget
Offsets passthrough inference spend
Policy checks
Context queries
Context indexes
Indexed storage
How inference bills
You pay the provider's rate with no markup, and the budget in your plan covers the first part of that spend. If you bring your own model-provider key, the provider bills you directly.
Inference rate model
You pay the upstream provider rate. Conjoin adds no markup. The included AI inference budget per plan offsets that spend, and the circuit breaker stops inference at the limit you set so a runaway prompt cannot drain the budget.
Inference throughput limits
Each plan sets its own ceiling on requests, input tokens, and concurrent inflight calls.
Plan
Build
Pro
Business
Enterprise
Requests per minute
Input tokens per minute
Concurrent requests
Inference capabilities
The table marks the plans that include each inference feature.
Plan
Build
Pro
Business
Enterprise
Multi-model routing
Best-of-N inference
Inference synthesis
Batch inference
Bring your own key
BYOK custom endpoint
Full workflow step library
Policy and safety
Conjoin AI checks every prompt and response against the topics that you block and the patterns that you write.
Custom policy capacity
Your plan sets how many custom patterns and topics you can keep active.
Plan
Build
Pro
Business
Enterprise
Custom patterns
Custom topics
Policy controls
The table shows what the policy engine can do on each plan.
Plan
Build
Pro
Business
Enterprise
Topic blocking
Response inspection
Log-only mode
Policy audit export
Context and retrieval
A context index lets your app answer from your own data. Your plan sets how many context queries you can make, how many indexes you can keep, and how much you can store in them.
The table marks the plans that include hybrid search, agentic context access, and query reranking.
Retrieval capabilities
Build
Pro
Business
Enterprise
Hybrid search
Agentic context access
Query reranking
Observability and retention
Traces, latency breakdowns, budget webhooks, and usage export help you audit AI traffic. Your plan sets which of them you get and how long Conjoin keeps the data.
Retention windows
Each plan sets how long Conjoin keeps observability traces and AI data.
Plan
Build
Pro
Business
Enterprise
Inference observability
AI data retention
Observability features
The table marks the plans that include each observability feature.
Plan
Build
Pro
Business
Enterprise
Per-request inference traces
Latency breakdowns
Budget threshold webhooks
Usage export
See your bill against real AI usage
Price your AI spend alongside every other Conjoin service. The calculator saves your numbers in the link, so anyone who opens it sees the same total.
Choose the plan that fits the stage of your app
Build will be free, while Pro and Business will add higher allowances and support. Tell us what you're building to join early access, or talk to our sales team about an Enterprise contract.