Pay only for the inference you use.

Every model has a clear input and output rate from the first request. No subscription. No seat fees. No commitment.

Create your Adola account Run a live demo

PRICING / 02

Included at no platform fee

Everything you need to build, test, and ship.

Go from API key to streamed response in minutes.

Keep your OpenAI client, choose a model per request, and watch tokens, latency, and spend from one workspace. Add billing to enable requests, then pay the published rate for the model you select.

Create your Adola account Copy the quickstart

OpenAI-compatible chat completions
Streaming responses
Explicit model selection
Project keys and usage controls
Per-request cost receipts
No recurring subscription, seat fee, or monthly minimum spend

PRICING / 03

Models and rates

4 models. 4 clear prices. One API.

Choose the model for each workload by ID. Adola runs that supported model or returns a clear error—never a silent substitute.

GPT-OSS 120B

OpenAI open-weight 120B MoE for capable reasoning, coding, and general-purpose work.

$0.14 input · $0.57 output
per 1M tokens
$0.001 successful-request minimum
131K-token context

gpt-oss-120b

Qwen3 30B

Fast multilingual MoE with strong tool use and agentic performance.

$0.11 input · $0.47 output
per 1M tokens
$0.001 successful-request minimum
41K-token context

qwen3-30b

Llama 3.3 70B

Meta's proven 70B instruction model for dependable general-purpose workloads.

$0.56 input · $0.75 output
per 1M tokens
$0.0015 successful-request minimum
131K-token context

llama-3.3-70b

DeepSeek V4 Flash

Fast, efficient reasoning and coding with a one-million-token context window.

$0.13 input · $0.26 output
per 1M tokens
$0.001 successful-request minimum
1M-token context

deepseek-v4-flash

PRICING / 04

Questions

Know the rate before every request.

Billing is required for API use. Compare models and published prices here, then monitor exact request costs in your workspace.

Do I need a card?

Yes. Add a payment method to enable API requests. Saving the card does not create a charge.

Do models cost different amounts?

Yes. Every model has its own published input and output price, so you can match capability and cost to the workload.

When do I pay?

Each successfully metered request costs the greater of its token subtotal or the model’s displayed request minimum.

Is there a subscription?

No. There is no monthly platform charge, seat fee, monthly minimum spend, or long-term commitment.

No platform subscription

Put a model in your product today.

Create an account, add a payment method, and pay only for successfully metered inference. No recurring subscription, seat fee, or monthly minimum spend.

Create your Adola account Copy the quickstart