ModelsModels / OpenAI
gpt-oss-120b
Gets 66 of every 100 tool calls right on the public test. Best at calling a single tool; it does not make several calls at once.
gpt-oss-120b is made by OpenAI. It takes text in and answers in text, with a context window of 131,072 tokens.
Start using gpt-oss-120b in 3 steps
Read the docs- Get an API keyFree to create. It works for every model in the catalogue.Get an API key
- Add creditsPay as you go. Credits work on every model.Add credits
- Point your client at TīrthaAny OpenAI-compatible client works. Use your key and this base URL, and name the model.base_url: "https://api.tirtha.ai/v1" api_key: "YOUR_TIRTHA_KEY" model: "catalog/gpt-oss-120b"
Price
The live price, read from billing as this page loads. You pay only for the tokens you use, from your credits.
At a glance
Tested on Tīrtha Bench
Tool calling (public BFCL v4 test)
Overall: 65.8%*. Picks the right tool and fills in the inputs correctly on 66 of every 100 tasks.
One tool 88% · Picking the right tool 86% · Several calls at once 0%
A task counts as right only if the model calls the right tool with the right inputs. 800 tasks. Run on Oct 29, 2026. *Measured by Spanda Works on third-party hosts, not through Tīrtha. Results through Tīrtha may differ.
Works best one call per step in our tests.
What it supports
- Tool callingYesAlso measured on Tīrtha Bench
- Text inYes
- ReasoningYesThinks before it answers
- StreamingYes
- Sampling settingsFixed todayThe maker publishes no default. Tīrtha fixes temperature at 0 today.
From the maker
- Model card: openai/gpt-oss-120b on Hugging Face
- Last updated by the maker: Aug 26, 2025
Facts here are read from the maker's published card; claims and benchmarks on that card are the maker's own, not measured by Tīrtha.
Licence
Weights: Apache-2.0. Read the licence.