GPT OSS 20B

• Type: Generation
• Capabilities: reasoning
• Cache read: $0.02 per 1M input tokens (0.65× standard input price). See prompt caching.

Overview

Meet gpt-oss-20b — for lower latency, and local or specialized use cases (21B parameters with 3.6B active parameters)

Reasoning efforts

• Supported: low, medium, high

See the reasoning effort guide for request examples.

Pricing

Priority	Input Tokens (per 1M)	Cache Read Tokens (per 1M)	Output Tokens (per 1M)
Realtime1	$0.03	$0.02	$0.13
Async	$0.02	$0.01	$0.10
Batch (24h)	$0.02	$0.01	$0.07

Playground

Open this model in the Playground.

Footnotes

1. Realtime availability is limited. Doubleword is primarily a batch API. ↩