Pricing
Unified Inference
Opaline uses a usage-based infrastructure model. Maintain platform credits and consume them as your applications access supported inference services — pay only for what your workloads consume.
| Model | Input $/1M | Cached $/1M | Output $/1M |
|---|---|---|---|
Qwen: Qwen3.8-27B | $0.2 | $0.05 | $2 |
Qwen: Qwen3.8-Flash-Next | $0.14 | $0.015 | $0.45 |
Z.AI: GLM-5.3-Flash | $0.07 | $0.015 | $0.24 |
Z.AI: GLM-5.3 | $1.1 | $0.23 | $3.5 |
DeepSeek: DeepSeek V4 Pro 0813 | $1.1 | $0.043 | $3.3 |
DeepSeek: DeepSeek V4 Lite | $0.1 | $0.02 | $0.4 |
Meta: Llama 4 Scout | $0.18 | $0.04 | $0.59 |
Mistral: Mistral Large 3 | $0.55 | $0.11 | $2.2 |
Moonshot: Kimi K2.5 | $0.6 | $0.12 | $2.5 |
MiniMax: MiniMax M2.5 | $0.3 | $0.06 | $1.2 |
Illustrative catalog — final pricing is announced with platform access.
What Is Metered
| Resource | What It Covers |
|---|---|
| Requests | Every call your application makes through the platform. |
| Input consumption | The input your workloads send to models. |
| Output consumption | The generation your workloads receive from models. |
| Model usage | Which models your applications use, and how often. |
| Spending | How your credits are consumed over time. |
| Application activity | Usage attributed to each of your applications. |
| Historical usage | Your consumption record across past periods. |
Credits and Platform Usage
The Opaline environment provides visibility into balances, usage, and expenditure, allowing developers to understand their ongoing AI infrastructure requirements.
The Opaline Token
The ecosystem incorporates a native token as part of its platform economy, intended to provide utility within eligible platform functions, participation mechanisms, and ecosystem incentives.
The token is not equity, ownership, or a claim on Opaline.
Pay for the intelligence you use. Nothing else.
Read the Details