API Catalog trial: can generated content be used in a commercial project, and what is the path to production?
Hello,
I am an independent creator building a small 3D animation pipeline (Blender) for an animated series for a teen audience (13+). I plan to publish it on YouTube and TikTok, with possible monetization. I use the NVIDIA API Catalog with a trial API key. NVIDIA Customer Care asked me to post my questions here.
Relevant API Trial Terms sections:
- 1.2 and 1.4: the trial is for testing and evaluation, not production use.
- 3.3 (iv): prompts and outputs may be used by NVIDIA to improve products.
- 4.2: Generated Content may not be distributed.
Models I intend to use:
- openai/gpt-oss-20b
- google/gemma-4-31b-it
- google/gemma-3-12b-it
- nv-mistralai/mistral-nemo-12b-instruct
- mistralai/mistral-7b-instruct-v0.3
Intended workflow:
- Now: only synthetic, invented test prompts (short dialogues, subtitle files, shot lists) to compare models. No personal data, no unpublished scripts, and the results are not distributed.
- Later: the chosen open-weight model is downloaded and run locally on my own GPU for production, under its own model license.
Questions:
- Is my current use (model comparison with synthetic prompts, results not distributed) within the trial terms?
- If a model is chosen through this evaluation and then run locally, is the local use governed only by that model’s own license, not by the API Trial Terms?
- Is there any way to use API Catalog generated content in a commercial production project (for example a paid plan or NVIDIA AI Enterprise), and where can I read those terms?
- What are the rate limits and credits for a trial account, and where can I see them?
- May trial outputs be used as training or fine-tuning data (distillation) for a local open-weight model that is later used in production? And may trial-generated helper code (for example small Python tools) be kept and used in my pipeline?
Thank you.