Groq (host) · LLMs & Coding

GPT OSS 20B

Catalog data · checked Nov 16, 2026

GPT OSS 20B is a language model from Groq (host), released in August 2025. It reads text and writes text, with a context window of 131.1K tokens and up to 65.5K output tokens per answer. It costs $0.075 per million input tokens and $0.3 per million output tokens. That makes it cheaper than 94% of the 185 paid language models we track.

Calculate my costCompare

Input price$0.075per 1M tokens
Output price$0.3per 1M tokens
Context window131.1Ktokens per request
Max output65.5Ktokens per answer
ReleasedAug 2025Aug 2025
Knowledge cutoffJun 2024training data ends
WeightsOpencan be self-hosted

What it can do

  • Tool calling
  • Structured output
  • Reasoning mode
  • Image input
  • File input
  • Audio input
  • Video input
  • Image output
  • Audio output
  • Prompt caching

More prices

Cached input (read)$0.0375 /1M

What it costs in practice

1,000 support replies1,000 chats with 800 input and 250 output tokens each$0.135
Summarize a 300-page reportabout 150,000 input tokens in one request, 2,000 tokens outtoo long
10,000 product descriptions300 input and 200 output tokens per text$0.825
One day of agentic codingabout 3M input and 150K output tokens across many calls$0.27

Calculated from the list prices above. Your bill depends on your real token counts. Use your own numbers →

Compare GPT OSS 20B with

Questions & answers

How much does GPT OSS 20B cost?

GPT OSS 20B costs $0.075 per 1 million input tokens and $0.3 per 1 million output tokens (USD, list price checked Nov 16, 2026).

How large is the context window of GPT OSS 20B?

GPT OSS 20B can process up to 131,072 tokens per request and return up to 65,536 tokens. As a rule of thumb, 1,000 tokens are about 750 English words.

Does GPT OSS 20B support tool calling?

Yes. The API accepts tool (function) definitions and also supports structured JSON output.

Can GPT OSS 20B read images?

No, this model accepts text only.