Openference plans
Pricing that scales with your agents
Deyin is free and open source — you only pay for model access through Openference, the unified LLM gateway that powers it.
Free
Everything you need to try agentic development.
10 requests/min
- 100 requests every 5 hours
- 350 requests per week
- Full open-source model catalog
- On-demand rates beyond the bundle
Lite
For daily drivers who delegate most of their work.
15 requests/min
- 400 requests every 5 hours
- 7,500 requests per week
- Full unified model catalog
- Community support
Pro
The sweet spot for serious daily agents.
30 requests/min
- 800 requests every 5 hours
- 13,000 requests per week
- Priority model availability
- Email support
Pro+
More headroom for heavy daily agent users.
30 requests/min
- 1,200 requests every 5 hours
- 19,500 requests per week
- Priority model availability
- Email support
Max
Large workloads, automations, and team throughput.
40 requests/min
- 1,600 requests every 5 hours
- 26,000 requests per week
- Priority access to new flagships
- Priority support
Max+
Fleet-scale autonomy with the highest limits.
40 requests/min
- 3,200 requests every 5 hours
- 52,000 requests per week
- Priority access to new flagships
- Highest autonomous throughput
How billing works
01
Create an Openference account — the free tier includes bundled requests that reset every 5 hours, no card required.
02
Sign in to Deyin with OAuth. Your plan authorizes every model call — no separate API keys to manage.
03
Upgrade or switch tiers anytime from Settings → Plans in Deyin. Upgrades apply immediately; downgrades at period end.
Every plan bundles requests that reset every 5 hours — go beyond the bundle at transparent on-demand rates, discounted on higher tiers. Annual billing saves roughly two months. Full plan details, regional pricing, and checkout live at openference.com/pricing. Prices shown in USD before regional adjustments.