Models/gpt-oss-20b

gpt-oss-20bActive · Cloud

OVHCloud AI Endpoints · Released Aug 2025 · Apache-2.0

gpt-oss-20b is an open-weight 21B parameter model released by OpenAI under the Apache 2.0 license. It uses a Mixture-of-Experts (MoE) architecture with 3.6B active parameters per forward pass, optimized for...

Context
131K
Max output
131K
Input
$0.04
Output
$0.15
Cached in
-
Latency
-

Capabilities

Core
Streaming
Function calling
JSON mode
Structured outputs
Multimodal
Vision
Image input
Audio
Image generation
Video
Advanced
Reasoning
Tool calling
MCP
Prompt caching
Batch API
Embeddings

Details

ReleasedAug 2025
LicenseApache-2.0
InputText
OutputText
Throughput-
Availability99.9%

Related models