gpt-oss-20b
General purposeToolsOpenAI · Released August 2025 · Knowledge cutoff June 2024
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...
It is part of OpenAI's GPT family, where Sol, Luna and Terra mark the current generation's capability tiers and Pro servings trade speed for deeper computation.
Facts and pricing
- Provider
- OpenAI
- Context window
- 131K tokens
- Input price
- 0.33 credits / 1M tokens ($0.030 raw)
- Output price
- 1.4 credits / 1M tokens ($0.13 raw)
- Vision (image input)
- No
- Tool calling
- Yes
- Extended reasoning
- No
Credits are what Zeplik bills: 1 credit = $0.10, computed from the raw provider rate with a 1.10x margin. Raw prices shown per 1M tokens.
What gpt-oss-20b is best for
- Balanced everyday work: writing, questions, brainstorming and summarization
- Tool use and agents: reliably calls functions, so it can search, run skills and drive workflows
Example prompts
Prompts that suit a general purpose model like gpt-oss-20b:
- Help me outline a talk about what I learned this year
- Compare these two paragraphs and tell me which is clearer and why
- Draft a polite reply declining this invitation
- Explain this concept with a concrete everyday analogy
OpenAI family
| Model | Context | Input cr/M | Output cr/M | Released |
|---|---|---|---|---|
| GPT-6 Astra | 1.1M | 110 | 550 | September 2026 |
| GPT-6 Astra Pro | 1.1M | 110 | 550 | September 2026 |
| GPT-5.6 Luna Pro | 1.1M | 2.2 | 13.2 | July 2026 |
| GPT-5.6 Luna Pro (batch) | 1.1M | 1.1 | 6.6 | July 2026 |
| GPT-5.6 Luna (batch) | 1.1M | 1.1 | 6.6 | July 2026 |
| GPT-5.6 Luna | 1.1M | 2.2 | 13.2 | July 2026 |
| GPT-5.6 Terra Pro (batch) | 1.1M | 11.0 | 66.0 | July 2026 |
| GPT-5.6 Terra Pro | 1.1M | 22.0 | 132 | July 2026 |
| GPT-5.6 Terra | 1.1M | 22.0 | 132 | July 2026 |
| GPT-5.6 Terra (batch) | 1.1M | 11.0 | 66.0 | July 2026 |
Frequently asked questions
- What is gpt-oss-20b?
- gpt-oss-20b is an AI model by OpenAI. 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...
- How much does gpt-oss-20b cost on Zeplik?
- Input tokens cost 0.33 credits per million and output tokens 1.4 credits per million (1 credit = $0.10; the raw provider rates are $0.030 and $0.13 per million). New accounts start with free credits.
- How long can a conversation with gpt-oss-20b be?
- gpt-oss-20b has a 131K-token context window (131,072 tokens), which covers the conversation plus any documents you attach.
- Does gpt-oss-20b support images and tools?
- gpt-oss-20b is text-only and supports tool calling.
Related models
GPT-6 Astra is OpenAI's flagship model for demanding end-to-end work. It is suited for advanced analysis, software engineering, deep research, scientific work, and document creation, with particular strengths in long-horizon...
GPT-6 Astra Pro is the same underlying model as [GPT-6 Astra](https://openrouter.ai/openai/gpt-6-astra), served with `reasoning.mode` set to `pro` for higher-quality responses on complex tasks. Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode
GPT-5.6 Luna Pro is the same underlying model as [GPT-5.6 Luna](https://openrouter.ai/openai/gpt-5.6-luna), served with `reasoning.mode` set to `pro` for higher-quality responses on complex tasks. Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode
GPT-5.6 Luna Pro is the same underlying model as [GPT-5.6 Luna](https://openrouter.ai/openai/gpt-5.6-luna), served with `reasoning.mode` set to `pro` for higher-quality responses on complex tasks. Learn more in OpenAI's docs: https://developers.openai.com/api/docs/guides/reasoning#reasoning-mode
GPT-5.6 Luna is a fast, cost-efficient model in OpenAI's GPT-5.6 series. It is suited for high-volume, latency-sensitive tasks such as chat, classification, and lightweight agentic workflows, providing capable reasoning for...
GPT-5.6 Luna is a fast, cost-efficient model in OpenAI's GPT-5.6 series. It is suited for high-volume, latency-sensitive tasks such as chat, classification, and lightweight agentic workflows, providing capable reasoning for...
More you can do with gpt-oss-20b
- 465 AI skillsReady-to-run expert methods for writing, data, code and more, each running on gpt-oss-20b or any model you pick.
- 997 integrationsConnect Gmail, Slack, GitHub, Notion and hundreds more so the assistant can work in your apps.
- Simple pricingPlus, Pro and Max plans with a monthly credit allowance that works across every model.