gpt-oss
gpt-oss is an OpenAI model family with 2 variants (120B and 20B) available in Quad Chat, with up to 128K tokens of context, tool calling and adjustable reasoning.
gpt-oss at a glance
| Lab | OpenAI |
|---|---|
| Available on | Free plan |
| Context window | Up to 128K tokens |
| Max output | Up to 128K tokens |
| Image input | No |
| Tool calling | Yes |
| Reasoning | Yes |
| Privacy lane | Private (Zero Data Retention) |
gpt-oss variants
| Model | Context | Plan | API list price per 1M tokens (in / out) |
|---|---|---|---|
| gpt-oss 120B | 128K | Free plan | $0.10 / $0.50 |
| gpt-oss 20B | 128K | Free plan | $0.03 / $0.14 |
How to use gpt-oss in Quad Chat
- Create a free account or sign in.
- Open the model picker and choose gpt-oss.
- Ask, attach files, or compare the answer with another model.
Frequently asked questions
Can I use gpt-oss for free?
Yes. gpt-oss is available on the Quad Chat Free plan, which needs no credit card and includes 100 monthly credits shared across every model. Larger requests use more credits, and paid plans from $8/month raise the allowance.
What is the context window of gpt-oss?
Context windows differ by variant: gpt-oss 120B 128K, gpt-oss 20B 128K tokens, according to the model catalog Quad Chat routes through. Files, pasted text and earlier messages all count toward the window.
Does gpt-oss support images, tools and reasoning?
gpt-oss works with text input only; pick a vision model in the same chat when you need to attach images. It supports tool calling. It can reason before it answers, with adjustable effort (low, medium and high). You can switch to another model in the same conversation at any time.
Is gpt-oss private in Quad Chat?
gpt-oss runs on the Quad Chat Private Lane, routed to providers under Zero Data Retention terms. Quad Chat does not train models on your conversations, encrypts chats in transit and at rest, and lets you export or delete them at any time.
How much does gpt-oss cost in Quad Chat?
Quad Chat plans have a flat monthly price and draw usage from shared credits, not a per-token bill. The model picker shows an estimate before you send. For reference, the provider's API list price is from $0.03 to $0.10 per million input tokens across variants.
Can I compare gpt-oss with other models?
Yes. Send one prompt to gpt-oss and one or two other models, such as GPT-6.1 Sol, GPT-6 and GPT-5.6, and compare the answers side by side in the same thread. You can also switch models mid-conversation without losing your files or context.
Related models
- GPT-6.1 Sol — OpenAI, Pro plan and up
- GPT-6 — OpenAI, Free plan for some variants; all from Pro
- GPT-5.6 — OpenAI, Free plan for some variants; all from Ultra
- GPT-5.5 — OpenAI, Ultra plan and up for some variants; all from Max
- Gemini 3.8 Flash — Google, Pro plan and up
- Claude Opus 5.5 — Anthropic, Ultra plan and up
- Kimi K3 — Moonshot AI, Free plan
- Muse Spark 1.3 — Meta, Free plan