Social Wire

Page 8 of 16 · newest first.

@sugaroverflow · CLI
Users are appreciating the small but thoughtful features of Devin CLI, such as persistent output and cute names for resumable sessions.
@thatguybg · Cognition
The AI industry is still in its early days, with many people hesitant to trust new technologies, much like the initial skepticism surrounding credit cards on the internet.
@Yuchenj_UW · Cognition
The AI coding model landscape is evolving rapidly, with many models converging in terms of performance, and smaller, cheaper, and open models emerging as viable alternatives.
@dabit3 · CLI
SWE-1.7 and GLM-5.2 models are free for the next month in Devin CLI and Devin Desktop, available to all subscription tiers.
@cognition · Cognition
Cognition announces the availability of Gemini 3.7 Flash in Devin, offering Claude Sonnet 5-level performance at less than half the cost
@devindesktop · CLI
Get ready to boost your productivity with Gemini 3.7 Flash, now available in Devin Desktop and Devin CLI at a 50% discount until August 27.
@GavinSBaker · Cloud
Get ready for the TomorrowXSummit, where Gavin Baker will interview Scott Wu, CEO of Cognition, about Devin's advancements in autonomous AI engineering and its impact on the industry.
@NancyZWang · Desktop
The latest episode of Zero-Shot Learning explores the future of software development, where AI coding agents fix bugs and vulnerabilities, with insights from Cognition's Jeff Wang and AnthropicAI's Richard Liu.
@aakashadesara · Cognition
A user recommends trying Cognition's SWE 1.7, citing its speed, affordability, and surprising performance
@CodeByPoonam · CLI
Gemini 3.7 Flash matches Claude Sonnet 5's performance on real coding tasks at less than half the cost, according to Cognition's FrontierCode 1.1 benchmark
@dabit3 · Cognition
Explore how X interactions impact your future feed with a real post and data-driven algorithm design, featuring Cognition and other industry leaders.
@mannupaaji · Cognition
A developer has successfully adapted to no-code editor views using Devin's startup program, coding with multiple agents and sessions
@rberrelleza · Cognition
A user praises DevinAI, feeling like they're working with a team of people rather than a single smart bot
@realbrucemartin · Cognition
Devin's SWE 1.7 Lightning achieves 1000 tokens/sec, outpacing GPT 5.6 Sol's 750/tok/s, according to a recent comparison
@cognition · Cognition
Grok 4.6 is now available in Devin, offering significant improvements over Grok 4.5 and surpassing GPT-5.6 Sol in performance
@devindesktop · CLI
Grok 4.6 is now accessible in both Devin Desktop and Devin CLI, expanding user options for leveraging its capabilities
@JustinGorya · CLI
Devin Desktop and DevinCLI now feature Grok 4.6 with enhanced reasoning capabilities, offering a significant boost in efficiency and intelligence.
@nateberkopec · Cloud
The shift towards fully remote, cloud-based environments is expected to continue, with potential implications for the way developers work and collaborate with AI agents.
@RebeccaTorrenc5 · Cognition
Cognition is reportedly in talks for a new funding round at a $40 billion valuation, with annualized revenue run rate approaching $1 billion, marking significant growth for the AI coding startup.
@ryancarson · Cloud
A developer shares their positive experience with DevinAI, citing the benefits of cloud-based engineering and the potential for a fully remote workflow in the future.
@bnistordev · Cognition
Devin is being used to build a Product Information Management system, handling tasks such as spinning up a VM, installing PrestaShop, and running end-to-end tests
@Sherveen · Cognition
Devin stands out from other harnesses for software engineering, offering a tier of its own, although it may be too expensive for some users
@jerryjliu0 · CLI
The future of FDE work may involve automated optimization processes, shifting the focus from implementation to defining goals and outcomes.
@odysseus0z · Cognition
Cognition's Frontier Code Bench highlights the challenges of building effective evaluations for code, with OSS maintainers spending significant time on this task.