Index / providers / novita-ai

Novita AI

GPU cloud and serverless LLM API with OpenAI-compatible endpoints, hosting classic uncensored roleplay fine-tunes alongside mainstream models.

Models4

Per-token price$0.050–$1.48/1M tok

Plans from—

Free tier—

novita.ai ↗Pending review

[ 01 ] Models & prices

BILLINGPay as you go per token. Card top-ups through Stripe, $10 minimum.

Model	Method	Context	In $/1M	Out $/1M	Per 1K chats
Sao10K/L3-8B-Stheno-v3.2Llama 3 8B · 8B	Fine-tuned	8K	$0.050	$0.050	$0.075
gryphe/mythomax-l2-13bLlama 2 13B · 13B	Fine-tuned	4K	$0.090	$0.090	$0.14
nousresearch/hermes-2-pro-llama-3-8bLlama 3 8B · 8B	Fine-tuned	8K	$0.14	$0.14	$0.21
sao10k/l3-70b-euryale-v2.1Llama 3 70B · 70B	Fine-tuned	8K	$1.48	$1.48	$2.22

Per 1K chats = cost of 1,000 messages of ~1,000 tokens in and ~500 out. Click a model to compare it across providers.All models →

[ 02 ] At a glance

Method: Fine-tuned
Access: Web chat + OpenAI-compatible API
Logging: No logs
Trains on prompts: No
Guarantee: Policy only
Jurisdiction: Delaware, US
Signup: Account required
Payments: Card
Data checked: Oct 11, 2026

PRIVACYTerms describe a "Zero Data Retention" policy where content isn't kept beyond what's needed to generate output (automated safety screening allowed), and content isn't used for training by default. Policy ↗

[ 03 ] Scores

Compliance

Benchmark pending

measured

Capability

Benchmark pending

measured

Privacy

48.0

from published policy & pricing

Truthfulness

Benchmark pending

measured

Cost

100.0

from published policy & pricing

Speed

Benchmark pending

measured

Composite: pending capability and compliance runs · method

[ 04 ] Notes & sources

Novita's own description of Euryale v2.1 calls it "the uncensored llama3 model". Also a Hugging Face Inference Provider.