Skip to main content

AI Coding Assistants · United Kingdom

Diffsift

AI code review that runs locally on your machine

  • Community votes0
  • DR by Ahrefs0 no change since last check

Ahrefs Domain Rating set in type 2 Nov 2026, reset 1 Nov 2026.

Diffsift turns pull requests from Azure DevOps and GitHub into structured review packs using your own local LLM, with no cloud, account, or telemetry.

The story so far

Screenshot of the Diffsift homepage
Homepage of Diffsift

Diffsift is an AI code review tool that runs entirely on your local machine or network. It watches your Azure DevOps and GitHub pull requests, generates structured review packs using a language model you control, and queues them in an inbox while you work. The tool integrates with LM Studio, Ollama, or any OpenAI-compatible endpoint.

The review process is simple: connect your repository with a personal access token, select a PR and review depth, and Diffsift builds a bounded context and runs your chosen model to produce findings with severity, confidence, and evidence from the actual diff. You copy the suggested comments you agree with and post them yourself—Diffsift never comments, approves, rejects, or merges anything.

Security and privacy are built in. By default, no code leaves your machine. If you configure an external provider like OpenAI, Diffsift warns you and requires explicit confirmation before each run. PATs and API keys are encrypted, and your review history is stored as plain Markdown on your disk. The app works offline with cryptographically signed license keys and never phones home.

Typeset features

  • Proactive review inbox with Windows notifications
  • Multi-pass review for large PRs
  • Evidence-linked findings
  • Review profiles for different focus areas
  • Review history stored locally
  • Supports local and external model providers
  • Azure DevOps and GitHub integration
  • Encrypted credential storage
  • Offline license activation

Pricing

Pricing model: Paid

Personal: $59 (one-time). Team: $119 per seat (minimum 2 seats, one-time). Renewal: 50% of original price ($29 personal / $59 per team seat) for 12 months of updates.

Diffsift metrics

Set in type for diffsift.dev: Ahrefs Domain Rating and Majestic link data, with the change since the last edition.
MetricValue
Domain Rating0 no change since last check
Trust Flow0 no change since last check
Citation Flow14 no change since last check
Referring domains2 no change since last check
External backlinks3 no change since last check

Ahrefs Domain Rating set in type 2 Nov 2026, reset 1 Nov 2026. Majestic set 1 Nov 2026, reset 1 Nov 2026. Link data typeset by Majestic. Trust Flow, Citation Flow and Topical Trust Flow are trademarks of Majestic-12 Ltd.

Diffsift domain colophon

Read the colophon
From the registry's public RDAP record, checked 3 Nov 2026, 12:50 UTC.
Domain age2 months
Registered
Expires
RegistrarCloudFlare, Inc.
DNSSECNot signed

Full colophon for diffsift.dev

Diffsift launch colophon

Measured after Diffsift went live on Zivaro Launch on 1 November 2026.

Domain Rating (Ahrefs)
0 no change since launch
0 at launch (1 Nov 2026), 0 on 2 Nov 2026
Referring domains (Majestic)
2
First measured 1 Nov 2026

Day 1 since launch. Next receipt: day 30, 1 Dec 2026. These are measurements, not proof of cause: many things move these numbers. Figures we have not measured are left out.

Biggest gains on Zivaro Launch

Questions and answers

Which models work?

Any model you can serve through LM Studio, Ollama, or an OpenAI-compatible endpoint works, including OpenAI, Azure OpenAI, Anthropic, and OpenRouter. Models in the 20B–35B range are optimal; larger models produce better reviews, smaller ones shallower findings. Tested local models include gpt-oss, Qwen 3.6, Gemma 4, DeepSeek-R1-0528, Muse Glimmer, Qwen 3.8, and Qwen 3.5.

What hardware do I need?

For local review, a GPU with 12 GB+ VRAM (or Apple-silicon machine with 32 GB+ unified memory) runs mid-size models at usable speeds. Smaller hardware works with smaller models but produces shallower findings. Diffsift itself is lightweight; the model host does the heavy lifting and can run on a different machine on your network.

Does any code leave my machine?

Not by default. With a local provider (LM Studio, Ollama, or any endpoint you host), nothing leaves your machine or network. If you configure an external provider like OpenAI, Diffsift warns you and requires explicit confirmation before every run, and every review pack records whether code may have left the machine.

Can it post comments automatically?

No. Diffsift generates copy-ready suggested comments, but you decide which ones are worth raising and post them yourself. Diffsift never comments on, approves, rejects, or merges a PR—you stay the reviewer of record.

Does it work with Azure DevOps Server / on-prem?

Diffsift targets Azure DevOps Services (dev.azure.com) and GitHub today. Azure DevOps Server support is on the roadmap — it uses the same REST API family, and on-prem shops are exactly who Diffsift is built for. Tell us you need it via the email list and it moves up.

Leave a review for Diffsift

No review of Diffsift set in type yet. Set the first.

Your rating

Reviews are free to set. None are paid for. Read the proofs

Readers' notes · Reprints · Share the page · Your company in print? Claim it · Report a misprint

More on this profile

Diffsift has a free listing. The owner can unlock the full profile with a paid launch for $5:

  • Founder story (locked)
  • Milestones (locked)
  • Demo video (locked)
  • Dofollow website link (locked)

More pages like this

  1. Number 1: y logoy

    A Mac coding agent app for Claude Code and Codex that reshapes itself live

    AI Coding Assistants

    DR 0 no change since last check 0 votes. Log in to vote for y in print
  2. Number 2: Amotions AI logoAmotions AI

    Emotionally intelligent AI sales coach for real-time conversation guidance

    AI Coding AssistantsUnited States

    DR 21 no change since last check 0 votes. Log in to vote for Amotions AI in print
  3. Number 3: Contextli logoContextli

    Context-aware voice dictation that automatically formats speech for each app

    AI Coding Assistants

    DR 51 no change since last check 0 votes. Log in to vote for Contextli in print
  4. Number 4: VocalCode logoVocalCode

    Push-to-talk voice input for coding with AI

    AI Coding Assistants

    DR 8 no change since last check 0 votes. Log in to vote for VocalCode in print
Help centre