Fitz 🔳 (@FitzGPT) header
Fitz 🔳 (@FitzGPT) profile picture

Anonim değilim.

Joined September 2025
Photo by Fitz 🔳 (@FitzGPT)Photo by Fitz 🔳 (@FitzGPT)Photo by Fitz 🔳 (@FitzGPT)Photo by Fitz 🔳 (@FitzGPT)Photo by Fitz 🔳 (@FitzGPT)Photo by Fitz 🔳 (@FitzGPT)Photo by Fitz 🔳 (@FitzGPT)Photo by Fitz 🔳 (@FitzGPT)Photo by Fitz 🔳 (@FitzGPT)Photo by Fitz 🔳 (@FitzGPT)
DeepSeek Harness içine Creator Mode eklemişler. 💥 Harness kendi eklentilerini kendisi yazıp kuruyor. Ne istediğini söylüyorsun, gerisini o hallediyor. Özel araç, çalışma panosu, hatta arayüzün tamamını yeniden tasarlatmak da mümkün. Altta Cordis var... Arayüzün kendisi dahil her şey eklenti olarak yazılmış. Yani özellik için geliştirici bekleme olayına hiç gerek yok istediğini doğrudan yazılıma söylüyorsun, o kuruyor.
Your Harness. You Define It. Meet Creator Mode in DeepSeek Harness. Turn ideas into plugins that you can use, from a design workspace to a sandbox game or a 3D LLM visualizer. What will you create?
1
4
29
1,923
Fitz 🔳 retweeted
Spiralens artık iphone da yayında! Hemen indirip deneyebilirsiniz!
Asked the cat to move a little left. Asked the sun to go up a bit. Nobody listened. Spiralens shows you where to move your phone instead: tap your target, follow the arrow, take the photo when it turns green. Now on iPhone, free: apps.apple.com/app/id6826378…
2
1
3
455
ÖZET: ZDR DİYE BİR ŞEY YOK.
1
313
MUAZZAM BİR OLAY ŞU.
Since I started using OpenCode I've wanted a native TUI. For the last week I've been daily driving a 1:1 port of the CLI, built on new OpenTUI bindings. It talks to the same TypeScript server.
282
Fitz 🔳 retweeted
Introducing Purrloader, the quiet way to save videos. For Mac, Windows and Linux! It's completely free and open-source. Paste. Purr. Saved. 🐱
3
2
1
19
1,817
Fitz 🔳 retweeted
Spiralens'i Rusya'nın büyük teknoloji sitelerinden Lifehacker.ru, ekim ayının en iyi Android uygulamaları listesine 1. sıradan koymuş. Başvurmadık, kimseye yazmadık. Yazıyı sitenin baş editörü yazmış; uygulamayı kendisi bulup seçmiş. Biz de uygulamaya bir anda Rusça kullanıcı gelmeye başlayınca peşine düştük ve bu yazıyı bulduk. Türkiye'den yaptığım bir uygulama için büyük an. lifehacker.ru/prilozheniya-a…
3
4
5
769
No daemon, no cron, no Node. 💥💥 Herdr App just dropped and I’ve already built on it. Here is why it hits different: 👇 Zero Background Footprint: No heavy runtimes or wasted cycles eating up your server resources. Pure Performance: Stripped down to the metal with a minimalist architecture built for speed. Buttery Smooth: A sleek 30fps native GPU spinner for instant, fluid visual feedback. More Herdr plugins dropping this week. Who knows, maybe one of them will finally save you from tab overload. 😉 Check out the repo:
less than 24 hours in: 30+ mods and 20+ layouts from you 🐏 this is what we believe in. herdr is your app. you don't wait for us to build a feature, you add it. not just themes. look 👇
1
4
462
Fitz 🔳 retweeted
‼️ Round 4 - Part 2 - DeepSeek API vs OpenCode Go vs OpenDesign vs Ollama API SWE Bench - Verified, 300 tasks. DeepSeek V4.1 Flash 1017. DeepSeek API - 283/300 Ollama Cloud - 284/300 OpenCode Go - 281/300 OpenDesign - 273/300 Interesting! What do you guys think?
🔔 Round 4, Part 1 - 100 SWE-Bench Verified Tasks Four Providers: DeepSeek API, OpenCode Go, OpenDesign, and Ollama API. One model, One reasoning: DeepSeek V4.1 Flash 1017 on Max. Results are VERY close. DeepSeek API - 95/100 OpenCode Go - 95/100 Ollama API - 95/100 OpenDesign - 94/100 Stay tuned for the more complete 300 task part 2 run soon!
28
6
8
193
23,748
KEL ADAM BİZİ SEVİYOR.
Operation Cheepseek Phase 4: $100 of usage for DeepSeek V4.1 Flash
1
4
596
⚡ @OpenRouter tip #6 Add the header X-OpenRouter-Cache: true and OpenRouter saves your entire response. Send the exact same request again and you get it back in about a second, for $0. This is different from prompt caching. Prompt caching only discounts the repeated part of your input, and the model still runs. This skips the model entirely. So if you're iterating during development you can turn this on to call the model without needing to pay for a new generation every time. The cache lasts 5 minutes by default, up to 24 hours with X-OpenRouter-Cache-TTL.
7
4
45
5,436
Keeping you in the loop. 5 OpenChamber releases in 11 days: - on a limit, auto-switch to another account or model - 160+ integrations, one-click MCP install - agents draw charts and mockups in chat - Gitea, Forgejo, Gerrit, stacked PRs - E2E-encrypted push with agent replies - @ a session for context
7
3
1
53
2,995
KİMİ 3.1 PRO İLE GERİ DÖNDÜ💥💥 Kodlama ve Blender performansı inanılmaz iyi. Haftasonu bi yakalım bakalım şu tokenleri neler olacak. Ve 4 kasımda açık kaynak olarak yayınlanacak ağırlıklar.
Introducing Kimi K3.1 Pro: our most capable 3T-class open-weight model. 🔹 1M-token context, 2.8T total / 104B active params (MoE), Multimodal. 🔹 Full-project coding: games in Unreal, Unity & Godot. 🔹 Excels at parametric 3D modeling in Blender & CAD. 🔹 More reliable at finance work: Accurate modeling with hallucination down by 65%. 🔹 Matching the strongest proprietary models across coding, agentic and knowledge benchmarks. Kimi K3.1 Pro and K3.1 Pro Highspeed (2–3× faster) are now live on Kimi.ai, Kimi Work, Kimi Code, and the Kimi API. Open weights by Nov 4, 2026. 🔗 API: platform.kimi.ai 🔗 Tech blog: kimi.ai/blog/kimi-k3-1-pro
1
1
12
22,467
Hi Friends! This time, instead of comparing different AI models, I compared the delivery performance of different AI Harness tools running the same model. We evaluated Pi, OpenCode, Claude Code, and Cursor using the same development prompt in the Port Crane Chess scenario. All tools used Sonnet 5.5 via OpenRouter, configured with medium effort. Task Overview: We asked the tools to develop a port simulation featuring autonomous crane control, physics-based cargo handling, dynamic task allocation, collision avoidance, and automatic task reassignment when a crane becomes unavailable. The evaluation focused on how quickly and cost-effectively each harness completed the development task, how easily the resulting software could be tested and inspected, and how effectively it reported its results. AI Harness Delivery Performance Scores /10 — Cost & Completion Time: 🥇 Pi — 8.46/10 — $2.634 — 17 minutes 🥈 Claude Code — 8.16/10 — $4.390 — 13 minutes 🥉 OpenCode — 7.68/10 — $2.370 — 25 minutes 🏅 Cursor — 7.24/10 — $3.780 — 19 minutes Key Findings: — Best balance between speed and cost — Pi. Ranking second in both metrics, Pi achieved the highest overall score. — Fastest completion — Claude Code. It completed the task in just 13 minutes, but also had the highest execution cost. — Lowest cost — OpenCode. However, it also recorded the longest completion time at 25 minutes. — Explicit testing and state inspection interfaces — Claude Code and Cursor. Both delivered implementations with named access points for externally stepping through execution and inspecting the current simulation state. — Restart behavior issue — OpenCode. Its implementation failed to clear the previous benchmark results after restarting. The cost–time trade-off was particularly interesting: — Pi completed the task 8 minutes faster than OpenCode for just $0.264 more. Compared with Cursor, Pi was 2 minutes faster and $1.146 cheaper. — These results demonstrate that the lowest execution cost does not necessarily translate into the highest overall efficiency. More importantly, this benchmark highlights a key observation: Using the same model, the same development prompt, and the same effort configuration does not guarantee identical completion times, execution costs, or output characteristics. — In this particular run, Pi delivered the most balanced performance, Claude Code was the fastest, and OpenCode was the most cost-efficient in absolute spending. Although Cursor's explicit testing interfaces were a positive aspect, it still fell behind Pi in both completion time and cost. Ultimately, the harness itself matters. Even when the underlying model remains identical, orchestration, execution workflows, and delivery behavior can significantly influence the final outcome.
12
11
6
142
16,894
-Pardon hesabı alabilir miyim? +Hesap ödendi! -Ne?! Kim ödedi? +Kel bir adam...
GLM 5.3 Flash is now free in OpenCode
4
2
182
3,380
Para verelim diyoruz istemiyorlar.
DeepSeek V4.1 Flash is now free in OpenCode
4
10
456
39,262
Fitz 🔳 retweeted
github'in kanayan yarasi size acilan pr ve issuelari rahatca gormek, tabii ki cobanov akli buna da bir care buldu
4
5
59
6,120
Operation Cheepseek Phase 3: $80 of usage for DeepSeek V4.1 Flash
1
13
902