DeepSeek logo Open weights · DeepSeek Model License

DeepSeek V3 Turbo

Novita AI's speed-optimized deployment of DeepSeek V3.

Jump to: Specs Pricing Self-hosting Similar models FAQ More from DeepSeek
Visit website Compare vs other LLMs →
Cheapest via Novita
Input
$0.40 /1M tokens
Output
$1.30 /1M tokens
Novita 1 provider

Key Specifications

Context window
64K tokens
Parameters
671B, 37B active
Inputs
Text
Capabilities Show details
Function calling Connect to external tools, APIs, and systems.

Hosted API pricing

Provider Input /1M tokens Output /1M tokens Cost at 10M in + 2M out
Novita logo Novita $0.40 $1.30 $6.60 View

Heads up: Base-tier, on-demand rates per 1M tokens; cached, batch and long-context tiers excluded. A provider may serve a shorter context or a quantized build than the creator's release. Verify before provisioning. More on how we price.

Estimated cost to self-host

DeepSeek V3 Turbo needs about 373 GB of GPU memory at 4-bit with 32K context. The cheapest rental that fits is 8× A100 at about $10,138 a month.

That costs the same as roughly 18B tokens a month on Novita's API. Self-hosting is more expensive below that volume.

Single node ≤ 8
Precision Memory Cheapest fit (32K context) Cost /mo Break-even vs API
4-bitINT4 / FP4
373 GB
Nvidia logo 8× A100
$10,138
18B tokens /mo
8-bitFP8 / INT8
674 GB
Nvidia logo 8× RTX PRO 6000
$12,672
23B tokens /mo
16-bitFP16 / BF16
1,346 GB
Nvidia logo 8× B300
$45,446
83B tokens /mo

Estimates based on median on-demand rates for Nvidia GPUs. Memory is weights plus KV cache for one request, using FP8 KV cache where supported and FP16/BF16 otherwise. Break-even assumes a 5:1 input-to-output ratio. No guarantee of runtime support, usable performance, or that a matching quantized build exists. How we estimate costs.

Similarly priced models

The models nearest DeepSeek V3 Turbo by blended rate, each at its own cheapest provider.

Model Blended /1M vs DeepSeek V3 Turbo
Google Cloud Gemini 3.1 Flash-Lite Google Cloud $0.4583 −17%
Alibaba Cloud Qwen3.6-Flash Alibaba Cloud $0.4583 −17%
Alibaba Cloud Qwen3-VL-235B-A22B Instruct Alibaba Cloud $0.50 −9%
DeepSeek DeepSeek V4 Pro DeepSeek $0.5233 −5%
OpenAI GPT-5 Mini OpenAI $0.5417 −2%
DeepSeek DeepSeek V3 Turbo This model DeepSeek $0.55
Alibaba Cloud Qwen3-VL-30B-A3B Thinking Alibaba Cloud $0.5667 +3%
Alibaba Cloud Qwen3-235B-A22B Thinking (2507) Alibaba Cloud $0.575 +5%
Alibaba Cloud Qwen3-Coder-480B-A35B Alibaba Cloud $0.575 +5%
OpenAI GPT-4.1 Mini OpenAI $0.60 +9%
Google Cloud Gemini 2.5 Flash Google Cloud $0.6667 +21%

Prices are USD per 1M tokens at each model's cheapest listed provider. Blended is the cost of 10M input plus 2M output tokens, spread over the 12M.

Frequently Asked Questions

What is DeepSeek V3 Turbo good for?

General chat and coding with tool calling, through Novita AI's own build of DeepSeek V3, which it says is tuned for speed.

When is DeepSeek V3 Turbo not a good fit?

Long-context work: its window is shorter than standard V3's. No structured output, and it is text only.

What is the cheapest way to run DeepSeek V3 Turbo?

Hosted, unless you push serious volume. Novita charges $0.40 in / $1.30 out per 1M tokens. The cheapest rental that fits is 8x A100 at $10,138 a month, which costs the same as about 18B tokens a month on that API.

Can I self-host DeepSeek V3 Turbo?

Yes. The weights are released under the DeepSeek Model License. At 4-bit it needs about 373 GB of GPU memory, which starts at roughly $10,138 a month on the cheapest rental that fits.

More from DeepSeek

Model Context Input /1M Output /1M
DeepSeek DeepSeek V4 Pro Open weights 1M $0.45 $0.89
DeepSeek DeepSeek V4 Flash Vision Exp 1M $0.44 $1.32
DeepSeek DeepSeek V4.1 Flash Open weights 1M $0.30 $1.20
DeepSeek DeepSeek V3 Open weights 128K $0.27 $1.12
DeepSeek DeepSeek V3.1 Open weights 131K $0.27 $1.00
DeepSeek DeepSeek V3.1 Terminus Open weights 131K $0.27 $1.00
DeepSeek DeepSeek R1 Distill Llama 70B Open weights 131K $0.80 $0.80
DeepSeek DeepSeek V3.2 Open weights 131K $0.25 $0.80