Cognition SWE-2 Flash
Cognition’s fast coding model, 3x the speed of SWE-2
Get picks like this daily. Today's launch numbers, the top launches, AI and tech news, and a daily opportunity pick, straight to your inbox.
Want to reach these readers? Sponsor an issue.
SWE-2 Flash is Cognition’s new coding model, built on Thinking Machines’ Inkling-Small with 276B total and 12B active parameters. Per Cognition, it generates about 200 tokens per second, 3x the speed of SWE-2 at one-fifth the price, and scores 20.0% on Terminal-Bench 4.0, up from 0.9% for the base model. Try it in Devin Desktop and Devin CLI, free through 11/30. Weights are not public yet.
PH · launched Nov 9, 2026 · 3 upvotes (top 7%) · 0 views
Niche: large language models · 191 launches, 4 in the last 7 days ▼ -33% · Explore this niche →
Category and tags
- Horizontal
- Dev tools
- Developer
- AI-native
- Commercial product
- Artificial Intelligence
Visit website ↗ Source listing ↗
More like this: all 191 launches in large language models →
Could you build this?
No This is a frontier coding model that was trained by Cognition on a large base model and proprietary training data. Vibe coding cannot produce a comparable model.
What it would actually take: Reproducing it requires a pretrained base model with hundreds of billions of parameters, large-scale agentic coding and terminal-task training data, and reinforcement-learning infrastructure. That means a research team with a large GPU cluster and substantial capital, not a product build.
Competitors
Other products that read as similar to this one : 86 launches clear the similarity bar, closest 8 shown.
Attention rank: #19 of 87 (itself plus its competitors, highest first, normalized so YC and Product Hunt are compared fairly).
Launched 332 days after the earliest competitor.
- Step 3.7 Flash · ph · 2026-05-30 · 183 upvotes · similarity 0.46
- Inkling 2 Small · ph · 2026-11-08 · 191 upvotes · similarity 0.43
- MiniMax M3.1 Flash · ph · 2026-10-26 · 171 upvotes · similarity 0.41
- Model Training Memory Simulator · hn · 2026-02-08 · 10 upvotes · similarity 0.38
- A decent text-to-image model trained for ~$20K (Apache 2.0 Weights) · hn · 2026-11-04 · 284 upvotes · similarity 0.38
- Speed Reading · ph · 2026-10-08 · 2 upvotes · similarity 0.37
- Naive-N0.5-Flash · github · 2026-09-27 · 142 upvotes · similarity 0.37
- Claude Cognitive · hn · 2025-12-30 · 15 upvotes · similarity 0.37
The niche, labels, vibe code rating and competitor comparison on this page are analysis by Nicheloom. If you use them, please credit nicheloom.com. How you may use our data.