Skip to main content

​
Pricing & Transparency

Openference offers the Minion plan ($20/month**) and the **Boss** plan (**$80/month) with rolling 5-hour (20%), weekly (50%), and monthly (100%) usage allowances, plus on-demand credits for pay-as-you-go usage or overage beyond your plan windows.

​
How plan allowance metering works

On Minion ($20/month) and **Boss** ($80/month), every request is metered against your 5-hour, weekly, and monthly allowances using the model’s standard catalog token rates (Input, Output, and Cached Read per 1M tokens), with no time-of-day or promotional discounts applied to plan allowance metering. When prompt caching is active, cached input tokens are metered at the model’s Cached Read rate.

​
Estimated requests (Minion)

Based on a typical coding turn of 10,000 input tokens and 100 output tokens (uncached) on the Minion ($20/month) plan:
  • Kimi K3 ($3.00/1M in, $15.00/1M out): ~380 / 5h · ~950 / week · ~1,900 / month
  • GLM-5.3 ($1.40/1M in, $4.40/1M out): ~830 / 5h · ~2,080 / week · ~4,160 / month
  • DeepSeek-V4-Pro-0813 ($1.32/1M in, $3.96/1M out): ~880 / 5h · ~2,210 / week · ~4,410 / month
  • Qwen3.8 27b ($0.45/1M in, $3.20/1M out): ~2,490 / 5h · ~6,220 / week · ~12,450 / month
  • DeepSeek-V4-Flash-0731 ($0.44/1M in, $1.32/1M out): ~2,650 / 5h · ~6,620 / week · ~13,240 / month
  • DeepSeek-V4.1-Flash ($0.30/1M in, $1.20/1M out): ~3,850 / 5h · ~9,620 / week · ~19,230 / month
  • GLM-5.3-Flash ($0.15/1M in, $0.50/1M out): ~7,740 / 5h · ~19,350 / week · ~38,710 / month
  • GLM-4.7-Flash ($0.06/1M in, $0.40/1M out): ~18,750 / 5h · ~46,880 / week · ~93,750 / month
  • Agnes-3.0-Flash ($0.05/1M in, $0.15/1M out): ~23,300 / 5h · ~58,250 / week · ~116,500 / month
  • Auto: — (dynamic routing)
  • Gemma 4 26B / GPT-OSS-120B / Llama 3.2 3B / MiniMax M3 / Nemotron-3-120B: Unlimited
See Model Costs & Usage Estimates, Minion Plan, Boss Plan, and Plans & Usage for the full table across every model in the catalog.

​
On-demand token rates

When you use on-demand credits (either without a subscription or after reaching a plan window limit), every model is billed per token at its published input, output, and cached-read rates in USD per million tokens. Rates are shown on:

​
Legacy request-ratio plans

Existing subscribers on retired legacy plans (Promo, Lite, Pro, Pro+, Max, and Max+) continue to use per-model request ratios until their current billing term ends — see Model Request Ratios (Legacy) for the full legacy ratio tables and runtime headers.