HFAI Cloud · model
GPT-OSS 120B
on HFAI Cloud.
Free to try, before you sign in.
What the catalogue publishes.
Read from the live model catalogue
at the moment this page was built.
| Value | As published |
|---|---|
| Identifier to send | gpt-oss:120b |
| Name in the catalogue | GPT-OSS 120B |
| Context window | 128,000 tokens (128K) |
| List price, per 1M input tokens | $0.15 |
| List price, per 1M output tokens | $0.60 |
| Free for anonymous callers | Yes |
Send it one request.
The same endpoint, key and request shape
as every other model in the catalogue.
The example below is complete apart from the key. The endpoint speaks the OpenAI chat-completions shape, so a client that already sends a model name, a message list and a bearer token needs no wrapper around it: point the base URL at https://api.heartflowai.net/v1, paste a key in place of the placeholder, and name this model with the identifier in the table above.
Streaming, tool calling and the error codes behave the same here as they do for the rest of the catalogue, so a harness you have already wired up keeps working when you change that one string.
curl https://api.heartflowai.net/v1/chat/completions \
-H "Authorization: Bearer hfai_live_…" \
-H "Content-Type: application/json" \
-d '{"model":"gpt-oss:120b","messages":[{"role":"user","content":"Hello HFAI"}]}'
Paste a key where the placeholder is. The request shape, the streaming rules and the error codes are documented in the API reference.
What is specific to it.
Read from the live catalogue,
plus one dated source.
GPT-OSS 120B is the free row here: $0.15 per 1M in, $0.60 per 1M out.
GPT-OSS 120B sits jointly 4th of 9 in and 4th of 9 out, priced $0.15 in and $0.60 out, the floor at $0.07 on GPT-OSS 20B.
GPT-OSS 120B charges $0.60 out for every $0.15 of input.
GPT-OSS 120B takes 128,000 tokens (128K) of context at $0.15 in and $0.60 out, with 3 rows above it.
GPT-OSS 120B needs no balance; the free tier covers this row alone.
GPT-OSS 120B comes from OpenAI under Apache-2.0: $0.15 in, $0.60 out.
The publisher recorded for GPT-OSS 120B is OpenAI.
GPT-OSS 120B is undercut on output by 3 rows, down to $0.30.
GPT-OSS 120B is requested as gpt-oss:120b, at $0.15 in, $0.60 out.
- Developer. gpt-oss:120b is built by OpenAI; checked · https://huggingface.co/openai/gpt-oss-120b
- Licence. gpt-oss:120b comes under the Apache-2.0 licence; checked · https://huggingface.co/openai/gpt-oss-120b
- Upstream model card. gpt-oss:120b has its card at https://huggingface.co/openai/gpt-oss-120b; checked · https://huggingface.co/openai/gpt-oss-120b
Every figure above comes from the 9 rows the live catalogue returned when the page was built, and is recomputed on every check, so it cannot quietly stop being true. The 3 sourced lines are the only facts here read from outside it. No speed, quality or benchmark figure is claimed on this page.
The rest of the catalogue.
This model shares one endpoint, one balance and one API key with every other model on offer, so moving between them costs a single string.