Novita AI

Novita AI (novita, Inference host; headquarters: Singapore) serves 20 models through the Autonomous Relay Agents API. Prices are the provider's list prices, per million tokens.

ModelModel idContextInputOutput
Llama 3 8B Stheno V3.2Sao10K/L3-8B-Stheno-v3.28,192$0.05 per M$0.05 per M
Baichuan M2 32Bbaichuan/baichuan-m2-32b131,072$0.07 per M$0.07 per M
CoBuddybaidu/cobuddy131,072$0.28 per M$1.13 per M
ERNIE 4.5 21B A3Bbaidu/ernie-4.5-21B-a3b120,000$0.07 per M$0.28 per M
DeepSeek OCR 2deepseek/deepseek-ocr-28,192$0.03 per M$0.03 per M
DeepSeek R1 Distill Qwen3 8B 0528deepseek/deepseek-r1-0528-qwen3-8b128,000$0.06 per M$0.09 per M
DeepSeek R1 Turbodeepseek/deepseek-r1-turbo64,000$0.70 per M$2.50 per M
Ling 3.0 Flash Santeinclusionai/ling-3.0-flash-sante262,144$0.075 per M$0.22 per M
Ling 3.0 Flash VLinclusionai/ling-3.0-flash-vl262,144$0.075 per M$0.22 per M
Macaron V1 Tallmindai/macaron-v1-tall262,144$0.2475 per M$1.43 per M
Macaron V1 Ventimindai/macaron-v1-venti1,048,576$0.825 per M$2.475 per M
MiniMax M2.7-highspeedminimax/minimax-m2.7-highspeed204,800$0.60 per M$2.40 per M
Kimi K2 Instructmoonshotai/kimi-k2-instruct131,072$0.57 per M$2.30 per M
Hermes 2 Pro Llama 3 8Bnousresearch/hermes-2-pro-llama-3-8b8,192$0.14 per M$0.14 per M
Bunnynovita/bunny262,144$0.00 per M$0.00 per M
PaddleOCR VLpaddlepaddle/paddleocr-vl16,384$0.02 per M$0.02 per M
Qwen MT Plusqwen/qwen-mt-plus16,384$0.25 per M$0.75 per M
Qwen3 Omni 30B A3B Thinkingqwen/qwen3-omni-30b-a3b-thinking65,536$0.25 per M$0.97 per M
Llama 3 70B Euryale V2.1sao10k/l3-70b-euryale-v2.18,192$1.48 per M$1.48 per M
AutoGLM Phone 9B Multilingualzai-org/autoglm-phone-9b-multilingual65,536$0.035 per M$0.138 per M