Our take
Novita AI is a Singapore-headquartered host with a very large catalogue of downloadable models and low entry-level pricing. It offers 163 downloadable-model options and a zero-retention setting, but compliance documentation and throughput data are thin.
Use this for exploratory access to a very broad open-weights catalogue — 163 offers to search. Pick it for cost-sensitive inference on small-to-mid parameter models where input pricing is well under a tenth of a cent per thousand tokens, or if you need a zero-retention option for prompt data. Skip it if your compliance review requires verified SOC 2 or a guaranteed EU endpoint, or if you need predictable throughput guarantees.
- Extremely large downloadable-model catalogue — 163 tracked offers.
- Very low entry pricing on small models.
- Does not train on prompts and offers a zero-retention option.
- Same model available at different throughput tiers with different pricing — you can pay more for higher speed.
- Compliance and regional documentation is thin: SOC 2 and EU endpoint are both unverified in our data.
- Throughput data is sparse and inconsistently reported — only 4 of 12 sample offers have it measured, and the same model can appear with or without throughput listed.
- Some models show high output pricing relative to input: output can cost several times the input rate.
Point your tools here
https://api.novita.ai/openaiModel IDs on this page are LLMap's own slugs, built for browsing and comparing — not what Novita AI's API expects. Its own model identifier lives in the docs above.
Privacy & data handling
These answers cover requests sent to Novita AI through OpenRouter, as OpenRouter records them, last read 4 hours ago. For Novita AI's own API we hold no answer: check its terms. SOC 2 and the data processing agreement below are hand-checked, not part of that sync.
Prompt retention: none.