OpenAI’s 21B-parameter open-weight Mixture-of-Experts (MoE) model, released under the Apache 2.0 license. Features 3.6B active parameters per forward pass, optimized for low-latency inference and deployability on consumer or single-GPU hardware. Trained in OpenAI’s Harmony response format, it supports reasoning level configuration, fine-tuning, and agentic capabilities such as function calling and structured outputs.
$0.002market price $0.004
$0.0015market price $0.003
$0.0085market price $0.017
Throughput is how fast the model writes (tokens per second, higher is better). Time to first token is how long before the answer starts (lower is better).
Share of requests served without a server error. Client errors (4xx) do not count.
Works with the OpenAI SDK pointed at our base URL. Every parameter is in the API reference.
Tokens processed per day on NagaAI.
OpenAI’s 21B-parameter open-weight Mixture-of-Experts (MoE) model, released under the Apache 2.0 license. Features 3.6B active parameters per forward pass, optimized for low-latency inference and deployability on consumer or single-GPU hardware.
On NagaAI, GPT OSS 20B costs $0.002 per 1M input tokens and $0.0085 per 1M output tokens. You pay only for what you use, with no subscription.
GPT OSS 20B reads up to 131.1k tokens in one request, counting the prompt and the answer together.
Yes. GPT OSS 20B can call your functions as tools and can answer in a JSON schema you give it.
It takes text and files as input and returns text.
Create an API key, copy the model ID from the top of this page and send requests with the OpenAI or Anthropic SDK pointed at NagaAI. The API section on this page has a ready example.
GPT-6.1 TerraOpenAIGPT-6.1 SolOpenAIGPT-6 LunaOpenAIGPT-6 SolOpenAIGPT-6 AstraOpenAIGPT 5.6 LunaOpenAI