Question 1

What is GPT OSS 20B?

Accepted Answer

GPT OSS 20B is a model from OpenAI, released on Aug 5, 2025. 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 lower-latency inference and deployability on consumer or single-GPU hardware. The model is trained in OpenAI’s Harmony response format and supports reasoning level configuration, fine-tuning, and agentic capabilities including function calling, tool use, and structured outputs.

Question 2

How much does GPT OSS 20B cost?

Accepted Answer

On the Respan gateway, input starts at $0.02 per 1M tokens on OpenRouter, and output starts at $0.09 per 1M tokens on OpenRouter, after gateway discounts. Averaged across 12 providers, it costs $0.05 per 1M input tokens and $0.20 per 1M output tokens.

Question 3

Which providers serve GPT OSS 20B on Respan?

Accepted Answer

12 providers serve GPT OSS 20B: OpenRouter, DeepInfra, Novita, Featherless, Parasail, TogetherAI, and 4 more. TogetherAI, Tera, and Groq require your own provider API key (BYOK).

Question 4

What is GPT OSS 20B's context window?

Accepted Answer

GPT OSS 20B accepts up to 131K tokens of context on its largest provider route.

Question 5

How do I use GPT OSS 20B with the Respan gateway?

Accepted Answer

Point any OpenAI-compatible SDK at https://api.respan.ai/api/ with your Respan API key and set the model to "gpt-oss-20b". To set it up with a coding agent, copy the agent setup prompt at the top of this page.

Question 6

How fast is GPT OSS 20B?

Accepted Answer

Over the last 30 days of requests through the Respan gateway, the fastest provider was Parasail at 114 tokens per second, and the lowest response latency was 0.8 seconds on Novita. Speed varies with request size and provider.

Question 7

Where does GPT OSS 20B rank on the Respan Index?

Accepted Answer

GPT OSS 20B scores 11.7 out of 100 on the Respan Index v1.1, ranking #133 of 138 models. By category: Coding 5.2, Agents 8.3 (estimated), Reasoning 10.1, Knowledge & language 7.3 (estimated), and Human preference 33.5. The Respan Index combines independent and lab-reported benchmark results into one score, using benchmark data collected by llmmetric.com.

Question 8

How does GPT OSS 20B score on benchmarks?

Accepted Answer

GPT OSS 20B has 8 published benchmark scores, including SWE-bench Verified 60.7%, GPQA Diamond 71.5%, and AIME 2025 89.2%. Each score links to where it was published; the data is collected by llmmetric.com.

Provider	Context	Input	Output
OpenRouter`openrouter/openai/gpt-oss-20b`	131K	$0.02/M	$0.09/M
DeepInfra`deepinfra/openai/gpt-oss-20b`	131K	$0.03/M	$0.14/M
NovitaSave 10%`novita/openai/gpt-oss-20b`	131K	$0.04/M	$0.14/M
Featherless`featherless_ai/openai/gpt-oss-20b`	131K	$0.04/M	$0.15/M
Parasail`parasail/parasail-gpt-oss-20b-fast-mem0`	131K	$0.04/M	$0.15/M
Parasail`parasail/openai/gpt-oss-20b`	131K	$0.04/M	$0.20/M
Parasail`parasail/parasail-gpt-oss-20b`	131K	$0.04/M	$0.20/M
TogetherAIBYOK only`together_ai/openai/gpt-oss-20b`	131K	$0.05/M	$0.20/M
TeraBYOK only`tera/openai/gpt-oss-20b`	131K	$0.07/M	$0.25/M
Bedrock`bedrock/openai.gpt-oss-20b-1:0`	131K	$0.07/M	$0.30/M
IO.net`ionet/openai/gpt-oss-20b`	64K	$0.07/M	$0.27/M
GroqBYOK only`groq/openai/gpt-oss-20b`	131K	$0.07/M	$0.30/M

GPT OSS 20B

Providers

Certified where it counts.

Pricing

Cost by volume

Throughput and latency

Output speed

Time to first token

Response latency

Benchmarks

Respan Indexv1.1

Coding2

Reasoning

Apps

GPT OSS 20B FAQ

Works with your entire stack

Built for AI agents.
Break less.
Ship more.

Math2

Arena1