Your cookie choices

We use necessary cookies to make this site work. With your permission, analytics cookies help us understand usage and marketing cookies support advertising.

You can reject non-essential cookies and still use the site. Change your choices anytime in Cookie settings.

Cookie settingsCookie policy

Novita AI review

Updated September 28, 2026 · DeployCue Editorial Team · Facts checked September 26, 2026

We earn a commission when you sign up through links on this site. Partner status can affect where a brand appears in our rankings. Partners with a live offer in your country come first, then every other brand by score.

Our verdict

Novita AI suits developers who want text, image, video and speech models plus GPU instances under a single account, with input prices from $0.02 to $3 and output prices from $0.03 to $15. The API is OpenAI-compatible, so existing client code keeps working when you move across.

Key facts

Best for
Chat, media generation and GPU instances under a single provider.
Input price from
$0.020 to $3↗
Output price from
$0.030 to $15↗
Models offered
DeepSeek V4 Pro, DeepSeek V4.1 Flash, DeepSeek R1, DeepSeek OCR 2, Llama 3.1 8B Instruct, Llama 4 Maverick Instruct, Qwen3.8 Flash, Qwen3.8 Max, Qwen3.5 397B A17B, ERNIE 4.5 21B A3B, Gemma 4 31B, GLM 5.3, Ling 3.0 Flash, Macaron V1 Venti, MiniMax M2.7, Kimi K3, Kimi K2.6, Mistral Nemo, OpenAI GPT OSS 120B, OpenAI GPT OSS 20B, NVIDIA Nemotron 3 Nano 30B A3B, Step 3.7 Flash, BAAI:BGE-M3, Kling v3.0 Pro, Wan 2.6, Fish Audio Text to Speech, SVD↗
Largest context window
1,048,576↗
Free credits
Selected models priced free ($0 input and output, e.g. Ling 3.0 Flash Fin, Ling 3.0 Flash Sante)↗
OpenAI-compatible API
Yes↗
Compliance certifications
SOC 2↗
Support channels
Email, Phone, Help-centre↗
Official website
novita.ai

GPU cloud

GPU models
4090, A100, L40, 3090↗
Billing granularity
Per second↗
Pricing model
Hourly billing

Serverless

Price per million requests
$5,000 to $16,000↗
Price per GB-second
$0.0000032↗
Maximum run time
3,600 to 14,400↗

Facts checked September 26, 2026, 8 sources

Apps and profiles

Elsewhere
github.compoe.com

Pricing

  • Novita AI Qwen3 Next 80B A3B (USD)

    input-tokens: $0.15 / million

  • Novita AI Qwen3 Next 80B A3B (USD)

    output-tokens: $1.50 / million

  • Novita AI Qwen3 Coder 480B (USD)

    input-tokens: $0.38 / million

  • Novita AI Qwen3 Coder 480B (USD)

    output-tokens: $1.55 / million

  • Novita AI Qwen3 235B A22B (USD)

    input-tokens: $0.09 / million

  • Novita AI Qwen3 235B A22B (USD)

    output-tokens: $0.58 / million

  • Novita AI Qwen 2.5 72B (USD)

    input-tokens: $0.38 / million

  • Novita AI Qwen 2.5 72B (USD)

    output-tokens: $0.40 / million

  • Novita AI Mistral NeMo (USD)

    input-tokens: $0.04 / million

  • Novita AI Mistral NeMo (USD)

    output-tokens: $0.17 / million

  • Novita AI MiniMax M2 (USD)

    input-tokens: $0.30 / million

  • Novita AI MiniMax M2 (USD)

    output-tokens: $1.20 / million

  • Novita AI Llama 3.3 70B (USD)

    input-tokens: $0.14 / million

  • Novita AI Llama 3.3 70B (USD)

    output-tokens: $0.40 / million

  • Novita AI Llama 3.1 8B (USD)

    input-tokens: $0.02 / million

  • Novita AI Llama 3.1 8B (USD)

    output-tokens: $0.05 / million

  • Novita AI Kimi K2 (USD)

    input-tokens: $0.57 / million

  • Novita AI Kimi K2 (USD)

    output-tokens: $2.30 / million

  • Novita AI gpt-oss-20b (USD)

    input-tokens: $0.04 / million

  • Novita AI gpt-oss-20b (USD)

    output-tokens: $0.15 / million

  • Novita AI gpt-oss-120b (USD)

    input-tokens: $0.05 / million

  • Novita AI gpt-oss-120b (USD)

    output-tokens: $0.25 / million

  • Novita AI GLM-4.6 (USD)

    input-tokens: $0.30 / million

  • Novita AI GLM-4.6 (USD)

    output-tokens: $0.90 / million

  • Novita AI GLM-4.5 (USD)

    input-tokens: $0.13 / million

  • Novita AI GLM-4.5 (USD)

    output-tokens: $0.85 / million

  • Novita AI Gemma 3 27B (USD)

    input-tokens: $0.12 / million

  • Novita AI Gemma 3 27B (USD)

    output-tokens: $0.20 / million

  • Novita AI DeepSeek V3.1 (USD)

    input-tokens: $0.27 / million

  • Novita AI DeepSeek V3.1 (USD)

    output-tokens: $1.00 / million

  • Novita AI DeepSeek V3 (USD)

    input-tokens: $0.89 / million

  • Novita AI DeepSeek V3 (USD)

    output-tokens: $0.89 / million

  • Novita AI DeepSeek R1 Distill (USD)

    input-tokens: $0.80 / million

  • Novita AI DeepSeek R1 Distill (USD)

    output-tokens: $0.80 / million

  • Novita AI DeepSeek R1 (USD)

    input-tokens: $0.70 / million

  • Novita AI DeepSeek R1 (USD)

    output-tokens: $2.50 / million

Pros and cons

Pros

  • Input prices run $0.02 to $3 and output prices $0.03 to $15, under the €0.07 category median
  • Largest context window reaches 1,048,576, past the 1,000,000 median
  • GPU instances on 4090, A100 and L40 bill per second and charge hourly
  • Several models are priced free on both input and output, so prototypes start at no cost

Cons

  • Compliance is limited to SOC 2, which may not clear a fixed audit checklist
  • Sandbox runs stop at 3,600 to 14,400, too short for long-lived sessions
  • Support runs on email, phone and help centre, and live chat is not among them
Novita AI

Ready to try Novita AI?

Scored 9.3 / 10 by DeployCue.

See alternatives

Who it suits

  • Developers on an OpenAI-compatible client who want to point at a different base URL
  • Teams running video or speech generation beside chat models
  • Small teams needing a GPU instance without a separate cloud contract

Who should look elsewhere

  • Azure AI Foundry buyers who want the catalogue held to a single vendor
  • Supascale AI readers who want a routing layer across many named providers
  • Salad users who only need pay-as-you-go GPU capacity

How Novita AI compares

Frequently asked questions

What does it cost to start on Novita AI?+

Input prices run $0.02 to $3 and output prices $0.03 to $15 for the cheapest models listed, while the top of both ranges sits well above the €0.07 category median. Every model carries its own rate, so budget from the entry figure.

Which models can I call through the API?+

The catalogue spans chat, embedding, image, video and speech under DeepSeek V4 Pro, DeepSeek V4.1 Flash and DeepSeek R1. DeepSeek, Qwen, Llama, GLM, Kimi and Mistral lines sit beside open-weight releases, with Kling and Wan for video and Fish Audio for speech.

Is there a free option to try first?+

Selected models are priced free, with nothing charged on input or output, and Ling 3.0 Flash Fin and Ling 3.0 Flash Sante are among them. Free tiers are common in the category, where 81.7% of providers include them.

Will my OpenAI client code work unchanged?+

The API is OpenAI-compatible, so existing client code can point at a different base URL instead of being rewritten. Compatibility is the norm in this category, with 100% of providers sharing the same shape.

How much context can a single request carry?+

The largest context window listed is 1,048,576, above the 1,000,000 median across the providers compared. That measure places it 11 among the brands ranked, and suits whole codebases in one call.

Can I rent GPU capacity from the same account?+

GPU instances cover 4090, A100 and L40, billed hourly at per second granularity, so short jobs stop paying when they end. Sandbox compute is priced at $0.0000032 per GB-second, with each run capped at 3,600 to 14,400.

What does serverless compute cost?+

Serverless request pricing runs $5,000 to $16,000 across the tiers listed, against a category median of €0.39. The upper end of that range sits far above the median, so the tier you land on matters more than the entry figure.

How do I reach support?+

Support runs on email, phone and help centre, with a help centre, email and a phone line listed. Live chat is not among those channels, so time-sensitive problems are worth raising by phone or email.

Sources

  • docs.novita.ai/guides/gpu-instance-pricing.md checked September 26, 2026

    Backs: Billing granularity, Pricing model

    “Billing duration is accurate to the second and settled hourly”

  • docs.novita.ai/guides/introduction checked September 26, 2026

    Backs: OpenAI-compatible API

    “Add hosted AI models to your application with an OpenAI-compatible API.”

  • novita.ai checked September 26, 2026

    Backs: Compliance certifications, Largest context window

    “AICPA SOC 2-zertifiziert”

  • novita.ai/docs/guides/sandbox-pricing.md checked September 26, 2026

    Backs: Price per GB-second

    “RAM (memory) billed per GiB per second: $0.0000032/GiB/s”

  • novita.ai/docs/guides/sandbox-quota-limit.md checked September 26, 2026

    Backs: Maximum run time

    “Maximum session duration: Free 1 hour, Paid 3 hours, Enterprise 4 hours by default”

  • novita.ai/gpu-instance/console checked September 26, 2026

    Backs: GPU models

    “4090, A100, L40, 3090 sind verfugbar”

  • novita.ai/legal/privacy-policy checked September 26, 2026

    Backs: Support channels

    “please contact us at: support@novita.ai ... Calling our toll-free number: 1-800-NOVITA-AI (1-800-668-4822)”

  • novita.ai/pricing checked September 26, 2026

    Backs: Free credits, Input price from, Models offered, Output price from, Price per million requests

    “Ling 3.0 Flash Fin ... Eingabe Kostenlos ... Ausgabe Kostenlos”

Compiled from 8 source pages, checked September 26, 2026. We did not open an account.

Novita AI

Novita AI

Chat, media generation and GPU instances under a single provider.

9.3/10See alternatives