Skip to content
latestai.eulatestai.eu
SF

SWE-2 Flash

by Cognition

Website

Cognition's fast, efficient coding model, built for sub-agents handling delegated coding tasks.

UnverifiedCodingLanguage ModelsModel

Updated EU status checked Last shown

SWE-2 Flash at a glance

What is SWE-2 Flash?
SWE-2 Flash is a model from Cognition. Cognition's fast, efficient coding model, built for sub-agents handling delegated coding tasks.
Who makes SWE-2 Flash?
SWE-2 Flash is made by Cognition.
Does SWE-2 Flash work in the EU?
Unverified. No explicit evidence found. Cognition's terms mention export control and sanctions compliance, but no supported-country list or EU restriction was provided. Last verified 14 Nov 2026.
How much does SWE-2 Flash cost?
Pricing unknown. No pricing in the provided material. Licence: Proprietary.
Where was SWE-2 Flash shown?
In 1 creator video by Matthew Berman. Most recently by Matthew Berman on 11 Nov 2026, at 16:07 in “Gemini Carbon, NVIDIA CEO AFFAIR, Fable 5.2, Meta Muse Ember, GPT Image 2.6, Minimax H3: AI NEWS”.

Sources: the vendor's own pages (linked above) and creator videos at the linked timestamps. EU status is information, not legal advice; see how we check.

About

SWE-2 Flash is a coding model from Cognition, the company behind Devin. It was announced on Cognition's site on 2026-11-04, following SWE-2 in September 2026. According to a YouTube creator summary, it is post-trained from Inkling Small and positioned as a fast, efficient model.

It is intended for use by sub-agents on delegated coding tasks, so a larger orchestrating model can hand well-scoped work to a cheaper, quicker one. It fits within Cognition's wider Devin product line, including Devin Desktop, the CLI and Windsurf.

Details such as benchmark results, pricing, context length and API access were not in the provided material. Cognition's site does show a FrontierCode benchmark, but no SWE-2 Flash scores were provided. Weights do not appear to be published.

Key features
  • Fast, efficient coding model
  • Designed for sub-agents on delegated coding tasks
  • Post-trained from Inkling Small (per creator report)
  • Part of Cognition's SWE model family used with Devin
1 moment

As seen on YouTube

Watch the moment: the exact second top creators show it. Videos load only when you press play, via youtube-nocookie.com.

Keep exploring

Similar AI tools

Closest neighbours by meaning, not just by category.

MC

MAI Code 1 Flash

Model · Microsoft

1

Microsoft's coding-focused language model, positioned as a faster, more token-efficient alternative to Claude Haiku 4.5.

UnverifiedCoding
Release date n/a
MA

MAI-Code-1.1-Flash

Model · Microsoft

1

Microsoft's coding-focused AI model in the MAI family, announced with few public details.

UnverifiedCoding
Release date n/a
DE

Devin

Agent · Cognition

2

Cognition's autonomous AI software engineer that runs parallel cloud agents to take tasks from ticket to merge-ready PR.

UnverifiedCoding
12 Mar 2024Freemium
MI

mini-SWE-agent

Library

1

Minimal open-source AI software engineering agent in about 100 lines of Python that uses only bash.

Available in EUPlatforms & Infra
28 Jun 2025Open source
G3

Gemini 3.5 Flash

Model · Google

5

Google's faster, lower-cost Flash model in the Gemini 3.5 family, aimed at agentic and coding workloads.

Available in EULanguage Models
May 2026Freemium
G3

Gemini 3.8 Flash

Model · Google

7

Google's fast, low-cost Gemini Flash model with strong agentic coding results, plus a Cyber variant for defensive security.

UnverifiedLanguage Models
Release date n/aAPI