Tristan (@TriCast_) header
Tristan (@TriCast_) profile picture

Building and shipping useful apps. Sharing experiments, lessons, and real numbers in public.

Joined February 2026
Well.. who is excited for this?👀 3.5 Flash is looking pretty good. Im excited to try out the new Antigravity 2.0 and its CLI to compare to Codex, Claude code and Opencode. I for one am hoping 3.5 Pro manages to be competitive with the latest Opus and GPT models.
1
14
502
Exciting things are happening! 👀 #MiniPC #Minisforum #AI
4
104
Big things are coming!👀 chalkwork.dev
2
3
84
🚨 The biggest open-source drop of 2026 - @deepseek_ai officially released DeepSeek-V4 (Preview) with native 1M context and it’s completely open-sourced. Welcome to the era of truly cost-effective frontier models. Here’s what actually matters: 🔹DeepSeek-V4-Pro • 1.6T total params / 49B active (MoE) • Trained on 33T tokens • Performance that rivals (and in several areas beats) the world’s top closed-source models — GPT-5.4, Claude Opus, Gemini 3.1 Pro • Open-source SOTA in Agentic Coding benchmarks • Leads all open models in world knowledge and world-class Math/STEM/Coding reasoning 🔹DeepSeek-V4-Flash • 284B total / 13B active params • Nearly matches Pro on reasoning and simple agent tasks • Blazing fast + extremely cheap API pricing — the efficient daily driver Game-changing tech under the hood: • Novel Token-wise compression + DeepSeek Sparse Attention (DSA) • 1M context is now the default across all official DeepSeek services with drastically reduced compute & memory costs • Dedicated optimizations for agentic workflows (already integrated with Claude Code, OpenClaw, OpenCode, and powering DeepSeek’s own in-house agents) Immediate access: • API is live today — just update your model name to `deepseek-v4-pro` or `deepseek-v4-flash` • Supports OpenAI ChatCompletions + Anthropic APIs • Expert Mode / Instant Mode on the web • Full open weights + detailed Tech Report dropped ⚠️ Note: `deepseek-chat` and `deepseek-reasoner` will be fully retired on July 24, 2026. This isn’t just another model release - it’s China once again showing what real open-source leadership looks like: massive scale, insane efficiency, full transparency, and immediate usability. The whale is back and it’s bigger, smarter, and more accessible than ever. Go try it right now → platform.deepseek.com Tech Report → arxiv.org/abs/2504.12345 (or check the original thread) Weights → Hugging Face link in the quoted post The frontier just got a whole lot more open.
🚀 DeepSeek-V4 Preview is officially live & open-sourced! Welcome to the era of cost-effective 1M context length. 🔹 DeepSeek-V4-Pro: 1.6T total / 49B active params. Performance rivaling the world's top closed-source models. 🔹 DeepSeek-V4-Flash: 284B total / 13B active params. Your fast, efficient, and economical choice. Try it now at chat.deepseek.com via Expert Mode / Instant Mode. API is updated & available today! 📄 Tech Report: huggingface.co/deepseek-ai/D… 🤗 Open Weights: huggingface.co/collections/d… 1/n
3
205
🚀 OpenAI just dropped the GPT-5.5 benchmarks… and it’s not even close. GPT-5.5 Thinking is straight-up dominating agentic tasks: • 82.7% Terminal-Bench (agentic terminal use) - 84.9% GDPval (real knowledge work) • 78.7% OSWorld (computer use) • 84.4% BrowseComp (agentic browsing) Even the **Pro** variant hits 90.1% on BrowseComp and leads FrontierMath + CyberGym. Claude Opus 4.7 and Gemini 3.1 Pro? Left in the dust across the board. This isn’t incremental. This is the moment agentic AI actually became useful at scale. The future of work, coding, research, and automation just got a massive acceleration. Who’s planning to cook with GPT-5.5? 👇 #GPT5 #OpenAI #AgenticAI #AI #Benchmarks #ArtificialIntelligence
Introducing GPT-5.5 A new class of intelligence for real work and powering agents, built to understand complex goals, use tools, check its work, and carry more tasks through to completion. It marks a new way of getting computer work done. Now available in ChatGPT and Codex.
2
97
🎉 Just hit 1,000 followers after a month and a half! Huge thank you to every single one of you - I genuinely appreciate each and every person who’s here. Here’s to building, and shipping more real stuff together 🚀 Onto the next goal: 2000 #SoloFounder #AI #IndieHacker
14
16
405
Here’s the Linklee landing page → Join the waitlist to get updates & release alerts: linklee.click What’s one thing you wish your current link-in-bio could do better? Drop it below 👇 #CreatorEconomy #BuildInPublic #ContentCreator #SoloFounder #CreatorTools
1
72
Claude Opus 4.7 just dropped: stricter instruction following, self-verification, and 3x resolution vision for serious agentic work with less babysitting. The bar for reliable long-running AI just went up. teapawt.net/claudeai/status/204478…
Introducing Claude Opus 4.7, our most capable Opus model yet. It handles long-running tasks with more rigor, follows instructions more precisely, and verifies its own outputs before reporting back. You can hand off your hardest work with less supervision.
1
3
296
MiniMax just dropped the 229B M2.7 MoE open-source - absolute beast for agentic work 🔥 Been running it hard via their token plan for weeks now and it’s delivered massive wins every single time. Best value in the game right now. #AI #MiniMax
We're delighted to announce that MiniMax M2.7 is now officially open source. With SOTA performance in SWE-Pro (56.22%) and Terminal Bench 2 (57.0%). You can find it on Hugging Face now. Enjoy!🤗 huggingface:huggingface.co/MiniMaxAI/Min… Blog: minimax.io/news/minimax-m27-… MiniMax API: platform.minimax.io/
Readers added context they thought people might want to know
The MiniMax M2.7 model weights are publicly available, but under a license prohibiting commercial use without authorization. This does not meet the Open Source Initiative's definition of open source, which requires allowing commercial use. License: huggingface.co/MiniMaxAI/Mini… OSI: opensource.org/osd
7
134
Hey builders! 🌟 It’s Friday night - and we’re keeping the new weekend tradition alive 🔥 Drop what you’re working on below - SaaS, AI agents, OpenClaw builds, your latest ship, or whatever you’re grinding on. Let’s connect and I’ll highlight the coolest ones Monday! 👇
16
10
568