OpenAI
Chat with GPT-OSS 120B online
OpenAI's open-weight reasoning model — Apache 2.0, tool-ready, and yours to self-host later.
131.072K tokens
Context
65,536 tokens
Max output
Fast
Speed
Straight from GPT-OSS 120B
Three prompts, three unedited answers, generated on September 2, 2026 through NinjaChat. Click the prompt to copy it and run it yourself.
Write a Python function that merges overlapping intervals, plus three short tests. Keep the whole answer under 40 lines of code.
def merge_intervals(intervals):
"""Merge overlapping intervals.
Args:
intervals: List of [start, end] pairs.
Returns:
List of merged intervals sorted by start.
"""
if not intervals:
return []
intervals = sorted(intervals, key=lambda x: x[0])
merged = [intervals[0]]
for cur in intervals[1:]:
prev = merged[-1]
if cur[0] <= prev[1]: # overlap
merged[-1] = [prev[0], max(prev[1], cur[1])]
else:
merged.append(cur)
return merged
# --- short tests -------------------------------------------------
assert merge_intervals([[1, 3], [2, 6], [8, 10], [15, 18]]) == [[1, 6], [8, 10], [15, 18]]
assert merge_intervals([[1, 4], [4, 5]]) == [[1, 5]]
assert merge_intervals([]) == []
print("All tests passed.")
Read: We ran the same 3 prompts on 12 AI models — Unedited answers, latency and price for every model, measured through the NinjaChat API in September 2026.
Served by Fireworks as gpt-oss-120b. Answers vary run to run; these are one take each, not the best of several.
What you’ll use it for
Coding help
Debug, explain and refactor with adjustable reasoning effort.
Agent-style tasks
Reliable tool calling from an open model.
Structured work
Extraction, classification and JSON at low cost.
Self-host later
Learn the model here, run the same weights yourself.
Prompts to steal
Tuned to this model — click any line to copy.
About GPT-OSS 120B
GPT-OSS 120B is the open-weight model OpenAI released in August 2025 under the Apache 2.0 license, the first open weights from the lab since GPT-2. It is a mixture-of-experts model with about 117 billion parameters and roughly 5 billion active per token, which is why it runs on a single 80GB GPU and why hosts can serve it fast and cheap. On NinjaChat it is included in every plan.
It is a reasoning model in the o-series tradition: you can ask for low, medium or high reasoning effort, it calls tools well, and it was trained with agentic use in mind. On the standard benchmarks it lands near OpenAI's small closed models of the same period, which for an open model you can inspect and self-host is the interesting part.
In daily use it is a strong default for coding help, structured tasks and anything where you want an open model with a familiar style. It is text only, so for screenshots and documents with images you will want one of the multimodal models beside it. And because the weights are public, what you learn here transfers if you later run it yourself.
How to use GPT-OSS 120B
From zero to your first result in under a minute.
01
Create a NinjaChat account and choose a plan
02
Open chat and pick GPT-OSS 120B from the model list
03
Ask for code, analysis or a structured answer
04
Say when you want it to think harder
05
Compare with the closed GPT-5.6 tiers in one click
Tips for better results
- Text only: use a multimodal model for images
- Raise the reasoning effort for math and logic
- It calls tools well; give it clear function descriptions
- The weights are Apache 2.0 if you want to self-host
Questions, answered.
Beyond chat — images & video
The same subscription covers the image studio and video generation.
GPT-OSS 120B, and 50 more.
One subscription covers every model on NinjaChat, GPT-OSS 120B included.
