Nothing to buy or switch on: it speeds up code search inside Devin Desktop, including on the free plan.
Kenning's take
What it does
SWE-grep and SWE-grep-mini are small models from Cognition trained to find relevant code quickly by running many searches in parallel. Cognition says they match frontier coding models at finding context while taking a fraction of the time. They power the Fast Context feature, first in Windsurf and now in Devin Desktop.
- Highly parallel code search
- SWE-grep and SWE-grep-mini versions
- Powers the Fast Context subagent
- Trained with multi-turn reinforcement learning
Who it suits. Developers using Devin Desktop or Windsurf who want faster code search inside the agent.
Key facts
- Category
- AI models
- Made by
- Cognition
- Released
- 16 October 2025
- How to get it
- built into Fast Context in Devin Desktop (formerly Windsurf); runs automatically when a query needs code search
- Works on
- Devin Desktop
- Free plan
- Devin's free plan includes Fast context, per the plan comparison
Pricing
| Plan | Price | Billed |
|---|---|---|
| Pro | $20 | month |
FAQ
When was SWE-grep released?
Cognition introduced SWE-grep and SWE-grep-mini on 16 October 2025.
How do I use SWE-grep?
It runs inside the Fast Context feature, which triggers automatically when a request needs code search.
What is SWE-grep for?
Finding the files and code an agent needs quickly, before it starts editing.
Facts from the maker's own pages and the sources listed. Checked 2026-10-11.
Similar tools
ACE-Step 1.5
Open-source music generation model that writes full songs with lyrics and runs on ordinary consumer GPUs.
AlphaGenome Atlas
Google DeepMind database of AI-predicted effects for every possible single-letter change in human DNA.
Amazon Nova 2
Amazon's second family of Nova AI models, covering reasoning, speech conversation and combined text-and-image generation.
Cisco Antares
Cisco's small open-weight models that find where known security vulnerabilities sit in a codebase, running locally.