DeepSeek V1 Author Open-Sources Personal Agent Project Linger
Tony Zhang, former DeepSeek V1 author and DeepSeek-VL project lead, unveiled Linger, an open-source personal agent that runs on the desktop rather than as an app. Its core is a local memory hub called Another Me that nightly distills email, calendar and chat into a personal wiki, scoring 95% on the LoCoMo long-term memory benchmark, while supporting third-party and locally optimized LLMs.
Related Entities
Full Report
Contents3 sections
Gathered, we are a blazing fire; scattered, we are a sky full of stars.
Just now, Tony Zhang, author of the original DeepSeek V1 and lead of the DeepSeek-VL project, officially unveiled his new startup project: Linger.
The direction is exactly the hottest thing right now: the personal Personal Agent.
X post video.mp4
The approach is also quite interesting. While peers are all building apps, they chose the computer first.
This means Linger can operate the web pages, inbox, and calendar already on your computer, with no need to wrestle with a pile of login credentials.
When it gets stuck somewhere, you can also take over directly.
X comment video.mp4
Another point: as a middle layer, Linger supports connecting to various third-party models.
And it has been specifically optimized for local LLMs.
Specifically, a single RTX 3060 graphics card with 12GB of VRAM can run a 125B-parameter Qwen3.8-Flash-Next, at a generation speed of 45.5 tokens/second.

Oh right, and one rather important point—
This system is open source...
As for why?
In the announcement post, Tony Zhang cited the history of the PC's rise:
In the 1960s, computing was centralized, until the PC put power into everyone's hands. Today, AI is trained in data centers and then rented to you in the form of tokens. We believe intelligence should be distributed. Your context, your data, and the intelligence that drives them should all be fully owned by you.
So today, they open-sourced Linger, a dedicated Agent that lives on your personal computer.

This also continues the fine tradition of the big fat whale 😭.
An AI that "takes initiative and can teach itself"
At the core of Linger is an Agent system called "Another Me."
It is first and foremost a memory hub.
Linger reads your inbox, calendar, chat records, and meetings locally, and every night automatically organizes this information into a personal knowledge base (Wiki) for all Agents to query.
In the LoCoMo benchmark, which measures long-term conversational memory, Linger's accuracy reached 95%, surpassing memory-augmented AI systems such as Mem0 and MemOS.

With this, Linger can work proactively more effectively.
It quietly observes your information stream and decides when to remind you based on the importance and urgency of events. If it receives a client email on Monday requesting a quote by Friday, it won't interrupt you immediately; instead, it waits until Wednesday morning to pop up a reminder.
But if a meeting time changes, it will notify you immediately.

This Agent also learns from your feedback, which Linger calls "implicit reinforcement."
For example, when it reminds you that an invoice is about to expire and you reply, "I received it last week, you should have told me then," the system records a clue.
If a similar situation arises with another invoice next time and you express the same preference again, the system automatically generates a rule: "Supplier invoice → remind immediately."

Conversely, if you show disinterest in "delivery reminders" twice in a row (for example, by ignoring them or replying "don't tell me"), it will also learn to stay silent in the future.
You can also give direct instructions to teach it all at once. For example, tell it: "Notify me immediately when a client replies," or "After 9 p.m., no need to tell me about meeting changes right away."
This continuous interaction and learning makes the Agent increasingly understand your work habits, turning it into a truly personalized assistant.
In addition, Linger can also:
- Operate the desktop and browser.
As mentioned at the beginning, Linger can directly use Chrome and various apps already logged in on your computer, with no reconfiguration needed.
- Build an Agent team.
This is the Super Agent principle that DeepSeek explained before.
When you issue a complex task, Another Me will invoke or create Agents with different functions to collaborate on the work.
And you only need to stay in this one Session with Linger.
X comment video.mp4
- Voice interaction.
This mainly serves office scenarios.
Linger can transcribe meetings in real time, and after the meeting you can directly ask the Agent about the meeting content and have it follow up on to-dos.
In short, it is a Personal Agent optimized for office scenarios.
However, some netizens did not quite understand this product positioning:
Why not develop based on the DeepSeek harness 🤣 What's the difference from ds-bot?
In response, Tony Zhang said:
Linger's focus is optimizing the model and harness together: a 12GB graphics card can run a 125B model locally, with no per-token fees, and data never leaves the machine. In terms of product positioning: the core of DS-Bot is Bot-to-Bot and Bot-to-group collaboration; the core of Linger is person-to-own-Agent and Agent-to-local-environment.

A 12GB graphics card running a 125B model
To let consumer-grade hardware run hundred-billion-scale large models, Linger's core technology is a MoE inference engine called PolyStrata.
Its principle can be understood as an intelligent tiered storage mechanism.
For a massive MoE (Mixture of Experts) model, PolyStrata places the most frequently used and most active "expert modules" in the fastest GPU VRAM, puts the less frequently used ones in memory, and places the least frequently used ones on the SSD.

When the model needs to invoke a particular expert, the system dynamically moves data between the GPU, memory, and SSD. This way, there is no need to cram the entire model—tens or even hundreds of GB—into limited VRAM.
The effect is immediate.
According to official data, a 125B-parameter Qwen3.8-Flash-Next model can achieve a generation speed of 45.5 tokens/second on a computer equipped with a 12GB RTX 3060 and 64GB of memory.
On the more powerful RTX 5090 (32GB), the speed soars to 179 tokens/second.
To make it easier for users to run local large models, Linger also created a "Model Autopilot" feature.
It automatically detects your hardware configuration (GPU, VRAM, memory), and then recommends and installs the highest-performing model for you from a combination of 49 language models, two inference engines, multiple quantization schemes, and different context lengths.

According to SimplexAI's scoring system, local devices with different configurations can achieve capabilities comparable to mainstream cloud-based large models:
- 8GB GPU + 32GB memory: score 35.4, roughly equivalent to Gemini 2.5 Pro.
- Mac 24GB unified memory: score 51.9, roughly equivalent to Claude Opus 4.5.
- 12GB GPU + 32GB memory: score 57.3, roughly equivalent to Claude Opus 4.6.
This means your gaming laptop or MacBook will have a truly powerful AI brain that is on standby 24/7, incurs no token costs, and keeps data entirely local.
DeepSeek-VL Lead Starts a New Venture
Behind Linger is a team with deep technical backgrounds, and the résumé of its founder, Tony Zhang, is especially impressive.
(After all, it's from the DeepSeek lineage…)

Tony Zhang completed his undergraduate studies at Zhejiang University, earned his PhD at the Hong Kong University of Science and Technology in 2019, and then joined Microsoft Research Asia (MSRA), where he long conducted research on computer vision and generative models and was an author of multiple well-known projects such as VQ-Diffusion and Rodin 3D digital humans.
In 2023, he joined DeepSeek as an AI scientist.
There, he not only participated in the development of the early language model DeepSeek LLM V1 series, but also served as the Project Lead for the vision-language large model DeepSeek-VL. This model has been highly praised for its ability to understand images, documents, and charts in real-world scenarios.
After leaving DeepSeek, Tony Zhang joined Zhejiang University as a faculty member, serving as a Hundred Talents Program researcher and doctoral supervisor, and founded Simplex AI.
Linger is Simplex AI's latest work.

And what about the inference engine that serves as Linger's technical core and was built specifically for local deployment?
According to Tony Zhang, it was built by an 18-year-old prodigy on the team.
18 years old? Just graduated from high school?