Devin

“First impressions of Devin AI, SWE-2 and Fusion on coding tasks from a real project. The good and the bad parts and why the CLI won me over.”

“But if they're running Claude Code, Codex, or Devin as separate CLI tools inside Cursor's terminal, Airport still fills a real gap Cursor doesn't natively cover — multi-tool, multi-terminal status tracking that isn't tied to Cursor's own agent brand.”

“(8-day-late) But I'm still farming it nonstop on both Devin desktop ide and CLI too...been great experience. Besides, the 20$ credit is preserved (and can be used for all other frontier models from Astra 6.0 to Fable 5.1) when using this post-train Kimi-K3 model https://cognition.com/blog/swe-2.”

“At the moment, the biggest bottleneck we’re facing is that some clients say Windsurf and others say Devin.”

“Should i try devin pro or copilot pro for luna only ? Who provide more usage?”

“I’ve been thinking a lot about how tools like Claude Code, Cursor, Devin-style agents, and similar AI coding assistants are quietly (or not so quietly) rewriting the software delivery lifecycle.”

“Get yourself a coding agent like Cursor, Claude, or Devin, install LangGraph and Svelte, and start building some use cases. It is the only real way to learn.”

“Cognition released SWE-2, a coding model post-trained on Moonshot’s Kimi K3. Strong numbers on several benchmarks, though performance drops on newer ones—classic overfitting signal to watch.”

“Instead of using Fable 5.1 or GPT-6 Astra for every step, Devin’s coding agent used them mainly for the hard thinking, then handed most of the execution to a much cheaper model. The result: – almost the same coding performance”

“But I am out spending her right now. Bc I’m just promoting Devin from my phone and reviewing code. Have spent $200 so far today.”

“@Steve_Yegge Try Devin Desktop.”

“Devin's Fusion might be great for coding but boy does it suck at design”

“Look at coding: Claude getting better didn’t make Devin less relevant. It made Devin better. And the more”

“I’ve run it in Droid, Cursor, and Devin. It’s the only open model that holds the full context no matter the task. Coherent all the way through.”

“"Benchmark" as in the disaster Devin made up to claim all their models are amazing. Let's see a dozen benchmarks like every other model publishes.”

“It's important that I usually use/used Devin for med. size tasks, so I'm not comparing the experience with Astra or Fable.”

“I used Windsurf for quite long and they're basically dead for now after they became Devin.Everything feels dull and they're always several features behind while Cursor is just killing it every other week.I'm back to VsCode plus Copilot Pro.”

“As others have mentioned, it's really matured a lot and at this point is one of the best cloud-hosted, team-managed coding agents, when factoring overall UX, testing and QA lifecycle via its sandboxes, and its ability to be controlled with an API.”

“Devin lacks many features and doesn't even have an exec mode. Other than they subsidize the subscription, the added value here is low.”

“As I said in a sibling thread on this page:> Devin CLI is easily the worst-in-class coding harness I've ever been subjected to using, rife with bugs (up to and including dropping answers to the question tool), so I can't say I'm exactly inspired to try anything else the company produces.”

“Lived experience is what matters and I don’t have anyone in my life (Devin shop) saying good things about SWE other than it’s free. Hope I’m wrong and it’s not so bad this time”

“Their ads in SF are pretty funny. “Remember Devin? It’s good now”. Okay, very self-aware, Cognition.”

“Friends who used it found the bad parts of an intern (tons of handholding, review required) but it didn’t learn from mistakes or add throughput.They seem to love a good overpromise.”

“It’s preferable to keep inference capacity available for users using main Devin products than openrouter atm. Might change in the future.”

“I'd consider if it I could access the model outside Devin. No lock in for me, thank you very much.”