Novita

Name: ANT
Author: Antbase

AI inference provider available on Antbase

OpenAI-compatible

About Novita

Novita is an AI model provider accessible through Antbase. Use intelligent routing to automatically select the best model from this provider based on your task requirements.

Live Performance Data

Intelligence Score

48.0

Total Requests

Success Rate

8103.0%

Avg Latency

3318ms

Models

Tokens Processed

1.4M

Novita Models on Antbase

Model	Score	Requests	Success Rate	Avg Latency	p95 Latency	Capabilities
Kimi K2.5	63.0	1	10000.0%	1085ms	1085ms
DeepSeek R1 0528	52.0	1	10000.0%	6425ms	6425ms
Minimax M2.1	50.0	0	0.0%	1000ms	2500ms	🔧
Qwen3 Coder Next	50.0	0	0.0%	1000ms	2500ms	🔧
AutoGLM-Phone-9B-Multilingual	50.0	0	0.0%	600ms	1500ms	👁
MiniMax-M2	50.0	0	0.0%	1000ms	2500ms	🧠🔧
GLM 4.6	50.0	0	0.0%	1500ms	3750ms	🧠🔧
MiniMax M2.5	50.0	0	0.0%	1000ms	2500ms	🧠🔧
XiaomiMiMo/MiMo-V2-Flash	50.0	0	0.0%	600ms	1500ms	🧠🔧
PaddleOCR-VL	50.0	0	0.0%	600ms	1500ms	👁
Deepseek V3.2 Exp	50.0	0	0.0%	1000ms	2500ms	🧠🔧
Qwen3 VL 235B A22B Thinking	50.0	0	0.0%	1500ms	3750ms	👁🧠🔧
GLM 4.6V	50.0	0	0.0%	1000ms	2500ms	👁🧠🔧
Kat Coder Pro	50.0	0	0.0%	1000ms	2500ms	🔧
Qwen3 Next 80B A3B Instruct	50.0	0	0.0%	1000ms	2500ms	🧠🔧
Qwen3 Next 80B A3B Thinking	50.0	0	0.0%	1000ms	2500ms	🧠🔧
DeepSeek-OCR	50.0	0	0.0%	600ms	1500ms	👁
Deepseek V3.1 Terminus	50.0	0	0.0%	1000ms	2500ms	🧠🔧
Qwen3 VL 235B A22B Instruct	50.0	0	0.0%	1000ms	2500ms	👁🔧
Qwen3 Max	50.0	0	0.0%	2000ms	5000ms	🧠🔧
DeepSeek V3.1	50.0	0	0.0%	1000ms	2500ms	🧠🔧
Kimi K2 0905	50.0	0	0.0%	1500ms	3750ms	🔧
Qwen3 Coder 480B A35B Instruct	50.0	0	0.0%	1000ms	2500ms	🔧
Qwen3 Coder 30b A3B Instruct	50.0	0	0.0%	600ms	1500ms	🔧
Qwen3.5-27B	50.0	0	0.0%	1000ms	2500ms	👁🧠🔧
Kimi K2 Instruct	50.0	0	0.0%	1500ms	3750ms	🔧
DeepSeek V3 0324	50.0	0	0.0%	1000ms	2500ms	🔧
GLM-4.5	50.0	0	0.0%	1500ms	3750ms	🧠🔧
MiniMax M2.5-highspeed	50.0	0	0.0%	1500ms	3750ms	🧠🔧
Qwen3.5-397B-A17B	50.0	0	0.0%	1500ms	3750ms	👁🧠🔧
Qwen3.5-35B-A3B	50.0	0	0.0%	1000ms	2500ms	👁🧠🔧
Llama 3.1 8B Instruct	50.0	0	0.0%	600ms	1500ms	🔧
Gemma3 12B	50.0	0	0.0%	600ms	1500ms	👁🔧
GLM 4.5V	50.0	0	0.0%	1500ms	3750ms	👁🧠🔧
OpenAI: GPT OSS 20B	50.0	0	0.0%	600ms	1500ms	👁🧠🔧
Qwen 2.5 72B Instruct	50.0	0	0.0%	1000ms	2500ms	🔧
Mistral Nemo	50.0	0	0.0%	600ms	1500ms	🔧
MiniMax M1	50.0	0	0.0%	1500ms	3750ms	🧠🔧
DeepSeek R1 Distill Qwen 32B	50.0	0	0.0%	1000ms	2500ms	🧠🔧
Llama 3 8B Instruct	50.0	0	0.0%	600ms	1500ms	🔧
Wizardlm 2 8x22B	50.0	0	0.0%	1500ms	3750ms
DeepSeek R1 0528 Qwen3 8B	50.0	0	0.0%	600ms	1500ms	🧠🔧
DeepSeek R1 Distill LLama 70B	50.0	0	0.0%	1500ms	3750ms	🧠
Llama3 70B Instruct	50.0	0	0.0%	1500ms	3750ms	🔧
Hermes 2 Pro Llama 3 8B	50.0	0	0.0%	1000ms	2500ms	🔧
Qwen2.5 VL 72B Instruct	50.0	0	0.0%	1500ms	3750ms	👁🔧
L3 70B Euryale V2.1	50.0	0	0.0%	2000ms	5000ms	🔧
ERNIE-4.5-21B-A3B-Thinking	50.0	0	0.0%	600ms	1500ms	🧠
Sao10k L3 8B Lunaris	50.0	0	0.0%	600ms	1500ms
BaiChuan M2 32B	50.0	0	0.0%	600ms	1500ms
ERNIE 4.5 VL 424B A47B	50.0	0	0.0%	1000ms	2500ms	👁🧠
ERNIE 4.5 300B A47B	50.0	0	0.0%	1000ms	2500ms
Deepseek Prover V2 671B	50.0	0	0.0%	1500ms	3750ms
Gemma 3 27B	50.0	0	0.0%	1000ms	2500ms	👁🔧
DeepSeek R1 (Turbo)	50.0	0	0.0%	1500ms	3750ms	🧠🔧
L3 8B Stheno V3.2	50.0	0	0.0%	600ms	1500ms
Mythomax L2 13B	50.0	0	0.0%	600ms	1500ms
ERNIE-4.5-VL-28B-A3B-Thinking	50.0	0	0.0%	1000ms	2500ms	👁🧠🔧
qwen/qwen3-vl-8b-instruct	50.0	0	0.0%	600ms	1500ms	👁🔧
zai-org/glm-4.5-air	50.0	0	0.0%	1000ms	2500ms	🧠🔧
qwen/qwen3-vl-30b-a3b-instruct	50.0	0	0.0%	1000ms	2500ms	👁🔧
qwen/qwen3-vl-30b-a3b-thinking	50.0	0	0.0%	1000ms	2500ms	👁🧠🔧
Qwen3 Omni 30B A3B Thinking	50.0	0	0.0%	400ms	1000ms	👁🧠🔧
Qwen3 Omni 30B A3B Instruct	50.0	0	0.0%	400ms	1000ms	👁🔧
Qwen MT Plus	50.0	0	0.0%	1000ms	2500ms	🔧
ERNIE 4.5 VL 28B A3B	50.0	0	0.0%	1000ms	2500ms	👁🧠🔧
ERNIE 4.5 21B A3B	50.0	0	0.0%	600ms	1500ms	🔧
Qwen3 8B	50.0	0	0.0%	600ms	1500ms	🔧
Qwen2.5 7B Instruct	50.0	0	0.0%	600ms	1500ms	🔧
Llama 3.2 3B Instruct	50.0	0	0.0%	600ms	1500ms	🔧
L31 70B Euryale V2.2	50.0	0	0.0%	2000ms	5000ms	🧠
DeepSeek R1 Distill Qwen 14B	50.0	0	0.0%	1000ms	2500ms	🧠🔧
DeepSeek V3 (Turbo)	50.0	0	0.0%	1000ms	2500ms	🔧
Qwen3 4B	50.0	0	0.0%	600ms	1500ms	🧠🔧
MiniMax M2.7	50.0	0	0.0%	1000ms	2500ms	🧠🔧
Llama 4 Maverick Instruct	50.0	0	0.0%	1000ms	2500ms	👁🔧
Llama 4 Scout Instruct	50.0	0	0.0%	1000ms	2500ms	👁🔧
Llama 3.3 70B Instruct	50.0	0	0.0%	1000ms	2500ms	🔧
Qwen3 235B A22b Thinking 2507	50.0	0	0.0%	1000ms	2500ms	🧠🔧
Qwen3 235B A22B Instruct 2507	50.0	0	0.0%	600ms	1500ms	🔧
Qwen3 235B A22B	50.0	0	0.0%	1000ms	2500ms	🧠🔧
Qwen3 32B	50.0	0	0.0%	600ms	1500ms	🧠🔧
Qwen3 30B A3B	50.0	0	0.0%	600ms	1500ms	🧠🔧
OpenAI GPT OSS 120B	50.0	0	0.0%	600ms	1500ms	👁🧠🔧
DeepSeek-OCR 2	50.0	0	0.0%	600ms	1500ms	👁
Qwen3.5 397B	48.0	29	10000.0%	8902ms	33484ms
DeepSeek V3.2	47.0	13	10000.0%	14102ms	29530ms
Qwen3.5 122B	45.0	1	10000.0%	18496ms	18496ms
GLM 5.3	45.0	2	10000.0%	12399ms	17654ms	🧠🔧
Kimi K2 Thinking	35.0	2	0.0%	385ms	394ms	🧠🔧
GLM-5	34.0	1	0.0%	537ms	537ms	🧠🔧
Qwen3.5-122B-A10B	34.0	1	0.0%	520ms	520ms	👁🧠🔧
GLM 5.3 Flash	34.0	1	0.0%	660ms	660ms	👁🧠🔧
GLM-4.7-Flash	33.0	1	0.0%	1191ms	1191ms	🧠🔧
DeepSeek R1 0528	33.0	1	0.0%	1216ms	1216ms	🧠🔧
Kimi K2.5	32.0	1	0.0%	1544ms	1544ms	👁🧠🔧
novita/qwen3.5-397b:free	15.0	2	0.0%	90780ms	90807ms
novita/deepseek-v3.2:free	15.0	1	0.0%	85437ms	85437ms

How to Use Novita on Antbase

All Novita models are available through Antbase's unified OpenAI-compatible API. You can either route directly to a specific Novita model or let ANT's intelligent router automatically select the best model for your task.

python

from openai import OpenAI

client = OpenAI(
    base_url="https://antbase.ai/v1",
    api_key="your-ant-api-key",
)

# Let ANT auto-route (may use Novita if optimal)
response = client.chat.completions.create(
    model="ant:auto",
    messages=[{"role": "user", "content": "Hello!"}],
)

Quickstart Guide Try in Playground

Why Use Novita Through Antbase?

•No separate account needed — access Novita models instantly through your Antbase API key.
•Intelligent routing — ANT automatically selects the best Novita model for each task, or falls over to another provider if Novita is slow or down.
•Same token prices — Antbase charges zero markup on Novita token prices. You pay exactly what Novita charges.
•Automatic failover — if Novita has an outage, your requests seamlessly route to an equivalent model on another provider.