ऐप डाउनलोड करेंऐप पाएँ

Deepseek

Choose a starting file from Deepseek, then see how to open it on your computer.

Which Deepseek model can I run on my computer?

It depends on your memory, runtime and chosen file. Compare the recorded sizes below and leave memory for other apps. A size estimate is only a planning aid. Open the exact file record, check its license and current hosted status, and look for a run receipt before downloading.

A model family groups related releases; variants can use different weights, licenses and file formats. RAM is working memory while a model runs; disk space stores its downloaded files. A run kit is a recorded setup, not proof that every family file has been tested.

Family catalog snapshot as of .

This family page links to individual model records. For example, open a listed model’s file evidence for filenames and recorded checksums, then inspect any run evidence. Page-copy provenance: the revision record names apps/web/lib/family-pages.mjs, apps/web/lib/family-starter.mjs, apps/web/lib/aeo-page.mjs and commit 8871bab84e1cf80aea1686a944ec335078a6c036. This identifies the text source; it is not a model test receipt. No family-wide run is implied.

Key facts

Family
Groups related releases; variants can differ in license, format and size.
Repository total size
Can include alternative files; it is not the memory needed for one selection.
Hosted status
Applies per model file record; unknown availability stays unconfirmed.
Run receipts
Apply only to the evidence shown for a specific file, not the whole family.

Evidence you can inspect (plain text)

  • This family page lists catalog snapshots for deepseek models with per-row file links.
  • Each table row links to /artifact/{id}/files for that model’s exact files.
  • Open /artifact/hf-model-qwen-qwen3-8b/files for model exact filenames, availability and any recorded SHA-256.
  • Open /evidence/artifacts/hf-model-qwen-qwen3-8b for run-readiness evidence tied to that record; absent receipts stay unmeasured.
  • Open /api/artifacts/hf-model-qwen-qwen3-8b/card.md for machine-readable source facts (not a runtime test).
  • No family-wide run is implied; inspect run evidence per model.

Unmeasured states are stated explicitly; this page does not invent run scores or compatibility guarantees.

Which date applies to which fact

  • Family catalog snapshot:
  • Committed page-copy sources:
  • Upstream model metadata: Dates on artifact records when shown; not implied by page-copy revision
  • Run receipts: Only on /evidence pages when recorded; otherwise unmeasured

Page-copy revision (Updated line above) identifies committed text sources, not upstream model edits or test outcomes.

Updated · Page-copy source revision 8871bab84e1cf80aea1686a944ec335078a6c036 · Source files: apps/web/lib/family-pages.mjs, apps/web/lib/family-starter.mjs, apps/web/lib/aeo-page.mjs · Full revision record

A model family: the original model plus smaller and modified versions made by others. A shared name doesn't promise the same quality.

Catalog snapshot: As of

Which Deepseek file should I start with?

Deepseek is a group of related models. Start with one chat file for your computer.

Recommended starting file · by size

deepseek llm 7B chat

deepseek-llm-7b-chat.Q4_K_M.gguf

File size
4.22 GB
Your memory
16 GB
License
unknown

A single chat-model file below 60% of your memory. Smaller files leave more room for other apps.

Not tested on your computer. Check the app's requirements. File size does not establish speed or writing quality.

How to open this fileView model files
Why this file?

We prefer a smaller chat model in the common Q4_K_M file type when one is recorded. The choice uses file size, not a performance test. Check current download availability on the model page.

Publisher and model: TheBloke/deepseek-llm-7B-chat-GGUF

I'm new: take me through the steps

All 88 model records in this family

हम विज़िट गिनने के लिए एक वैकल्पिक कुकी इस्तेमाल करते हैं।

निजता