DeepSeek’s flagship DeepSeek-Prover-V2–671B model is now live on Novita AI’s Model API platform! For a limited time, all new users can claim $10 in free credits to explore and build with this state-of-the-art automated theorem proving model.

Here’s the Current DeepSeek-Prover-V2–671B Pricing on Novita AI:

DeepSeek-Prover-V2–671B: 0.7 $ / M input tokens, 2.5$/ M output tokens

Power your mathematical workflows, research, and educational tools with one of the most advanced formal theorem-proving models ever released — now just an API call away.

What is the DeepSeek-Prover Model Series?

The DeepSeek-Prover model series is a state-of-the-art family of large language models (LLMs) designed to tackle the challenge of automated theorem proving in mathematics. Developed by DeepSeek-AI, these models are specifically optimized for generating, verifying, and reasoning about formal mathematical proofs — especially within the Lean 4 proof assistant environment.

Why Was DeepSeek-Prover Created?

While modern LLMs like GPT-4 are impressive at mathematical reasoning, they struggle with formal proof tasks due to a lack of high-quality, large-scale training data in formal languages. DeepSeek-Prover solves this by auto-generating millions of synthetic Lean 4 proofs from math competition problems and fine-tuning models on this unique dataset.

How Does DeepSeek-Prover Work?

The DeepSeek-Prover pipeline includes:

• Autoformalization: Translating natural language math problems into formal Lean 4 statements using an LLM.

• Quality Filtering: Filtering statements for logical correctness and mathematical interest through model scoring and hypothesis rejection.

• Proof Generation & Verification: Attempting to prove filtered statements, and keeping only those with correct, verifiable proofs.

• Iterative Training: Repeating this process, each time using the new high-quality proof data to further improve the model.

Through this approach, the DeepSeek-Prover series produces models that continuously improve and are supported by one of the world’s largest collections of formal mathematics proofs.

Highlights and Milestones from Earlier DeepSeek-Prover Models

• Scale: Over 8 million formal Lean 4 statements with proofs were generated.

• Technical Foundation: Based on large transformer models, such as DeepSeekMath 7B, pre-trained on extensive mathematical data and further fine-tuned on synthetic proofs.

• State-of-the-Art Performance: Achieves top results on formal theorem proving benchmarks (e.g., miniF2F, FIMO), outperforming GPT-4 and other specialized models.

What Is DeepSeek-Prover-V2–671B?

DeepSeek-Prover-V2–671B is a flagship model in the DeepSeek-Prover series. Here’s what sets it apart:

• Massive Scale: With around 671 billion parameters, this is one of the largest models designed for formal mathematics and theorem proving.

• FP8 Precision: DeepSeek-Prover-V2–671B is originally trained and released in FP8 (floating point 8-bit) precision, a modern format that enables efficient computation and faster model serving with lower memory usage, while maintaining high performance.

• Advanced Capability: The model uses state-of-the-art transformer architecture and supports modern tensor types like BF16, FP8_E4M3, and F32, making it suitable for high-performance research and production environments.

• Purpose-Built for Theorem Proving: Trained on massive, high-quality synthetic proof data, this model excels at complex mathematical reasoning and formal Lean 4 proof generation.

How to Access DeepSeek-Prover-V2–671B on Novita AI

Getting started with DeepSeek-Prover-V2–671B is fast, simple, and risk-free on Novita AI. Thanks to the Referral Program, you’ll receive $10 in free credits — enough to fully explore DeepSeek-Prover-V2–671B’s power, build prototypes, and even launch your first use case without any upfront cost.

Use the Playground (No Coding Required)

• Instant Access: Sign up, claim your free credits, and start experimenting with DeepSeek-Prover-V2–671B and other top models in seconds.

• Interactive UI: Test prompts, chain-of-thought reasoning, and visualize results in real time.

• Model Comparison: Effortlessly switch between DeepSeek-Prover-V2–671B, Llama 4, DeepSeek, and more to find the perfect fit for your needs.

Integrate via API (For Developers)

Seamlessly connect DeepSeek-Prover-V2–671B to your applications, workflows, or chatbots with Novita AI’s unified REST API — no need to manage model weights or infrastructure. Novita AI offers multi-language SDKs (Python, Node.js, cURL, and more) and advanced parameter controls for power users.

Option 1: Direct API Integration (Python Example)

To get started, simply use the code snippet below:

from openai import OpenAI

client = OpenAI(
    base_url="https://api.novita.ai/v3/openai",
    api_key="<YOUR Novita AI API Key>",
)

model = "deepseek/deepseek-prover-v2-671b"
stream = True # or False
max_tokens = 2048
system_content = """Be a helpful assistant"""
temperature = 1
top_p = 1
min_p = 0
top_k = 50
presence_penalty = 0
frequency_penalty = 0
repetition_penalty = 1
response_format = { "type": "text" }

chat_completion_res = client.chat.completions.create(
    model=model,
    messages=[
        {
            "role": "system",
            "content": system_content,
        },
        {
            "role": "user",
            "content": "Hi there!",
        }
    ],
    stream=stream,
    max_tokens=max_tokens,
    temperature=temperature,
    top_p=top_p,
    presence_penalty=presence_penalty,
    frequency_penalty=frequency_penalty,
    response_format=response_format,
    extra_body={
      "top_k": top_k,
      "repetition_penalty": repetition_penalty,
      "min_p": min_p
    }
  )

if stream:
    for chunk in chat_completion_res:
        print(chunk.choices[0].delta.content or "", end="")
else:
    print(chat_completion_res.choices[0].message.content)

Key Features:

• Unified endpoint:/v3/openai supports OpenAI’s Chat Completions API format.

• Flexible controls: Adjust temperature, top-p, penalties, and more for tailored results.

• Streaming & batching: Choose your preferred response mode.

Option 2: Multi-Agent Workflows with OpenAI Agents SDK

Build advanced multi-agent systems by integrating Novita AI with the OpenAI Agents SDK:

• Plug-and-play: Use Novita AI’s LLMs in any OpenAI Agents workflow.

• Supports handoffs, routing, and tool use: Design agents that can delegate, triage, or run functions, all powered by Novita AI’s models.

• Python integration: Simply point the SDK to Novita’s endpoint (https://api.novita.ai/v3/openai) and use your API key.

Connect DeepSeek-Prover-V2–671B API on Third-Party Platforms

• Hugging Face: Use DeepSeek-Prover-V2–671B in Spaces, pipelines, or with the Transformers library via Novita AI endpoints.

• Agent & Orchestration Frameworks: Easily connect Novita AI with partner platforms like Continue, AnythingLLM,LangChain, Dify and Langflow through official connectors and step-by-step integration guides.

• OpenAI-Compatible API: Enjoy hassle-free migration and integration with tools such as Cline and Cursor, designed for the OpenAI API standard.

Conclusion

DeepSeek-Prover-V2–671B brings world-class automated theorem proving directly to your fingertips — whether you’re a researcher, developer, or educator. Ready to try it? Sign up on Novita AI, claim your free credits, and experience the future of formal mathematics and AI-driven proof generation today!

About Novita AI

Novita AI is an AI cloud platform that offers developers an easy way to deploy AI models using our simple API, while also providing an affordable and reliable GPU cloud for building and scaling.