Deepinfer · Office #48,603

What it is

Calls itself “inference engine”.

• Calls itself: inference engine

“…inference engine”

Where it puts itself

Rules out “Not a hosted API or an OpenAI-compatible server.”. Publishes a comparison page.

• Says it will not: Not a hosted API or an OpenAI-compatible server.

“Not a hosted API or an OpenAI-compatible server.”

“…https://deepinfer.net/blog/deepinfer-vs-vllm/”

Who it sells to

Sells to developers.

• Sells to: developers

“Post-train DeepSeek-V4.1-Flash or another frontier base on your industry data, then serve it on Mint Recursive.”

How it runs

Runs on Linux. Runs locally and works offline. No API. Open source. AI is the core of the product.

• Runs on: Linux
• Runs where: on the device, offline
• API: no
• Open source: yes
• AI role: core to the product

“Hardware SM100 / Ascend 950 Linux only”

“Scope Offline engine not a hosted API”

What it works with

Names DeepSeek-V4.1-Flash, LMDeploy and 4 more on its pages.

• Names: DeepSeek-V4.1-Flash · LMDeploy · SGLang · +3

“DeepInfer is DeepSeek's open-source inference framework for serving DeepSeek-V4.1-Flash at scale.”

“DeepInfer vs LMDeploy”

What it offers as proof

Live, no waitlist. Publishes docs and a blog.

“DeepInfer is an offline inference engine, not a hosted API or a model; the release does not include the KV cache system, prefill/decode disaggregation, or data-parallel load balancing needed for a complete online service.”

“Official Q&A -> https://github.com/deepseek-ai/DeepInfer/blob/main/doc/inference_deployment_qa.md”