One stack for all AI. Run any model, at the lowest price.

Joined March 2024
كل هالتنقل بين أدوات الذكاء الاصطناعي ..ليه ما تخليها تشتغل مع بعض؟ 👀 أنا أستخدم @runware لربط توليد الصور والفيديو والصوت بسير عملي في الموشن جرافيك، من خلال واجهة برمجية وحدة #Runware #موشن_جرافيك #الذكاء_الاصطناعي في المقطع أبسط لك الفكرة: كيف تطلب الأصل البصري، تستقبل النتيجة، وتدخلها في مشروعك عشان تبدأ التركيب والتحريك
Nano Banana 2.1 is on Runware 🍌 - Sharper composition and subject consistency - Up to 14 reference images and 10 reference videos - Up to 4K output across 14 aspect ratios - 3 thinking levels
15
105
6,695
you don't have to describe a pose to Ideogram 4.5
1
3
143
Runware retweeted
Astra + @runware MCP edited this for me
🤖 Made with AI
2
5
248
something I like about Gemini Omni Flash 1.1 is the video extension feature, works so nicely!
1
3
708
Runware retweeted
Wan 3.0 supports audio references, so obviously I had to make a music video. generated using audio+prompt+image of a singer using @runware
2
4
257
Runware retweeted
One of the coolest prompts for Seedance 2.5 (30s) ! Prompt ⤵️
🤖 Made with AI
31
27
6
320
22,654
threw the newest video models at one 10-second one-shot: a sword duel in a bamboo grove Wan 3.0, LTX 2.5 Pro, Flux 3, Minimax H3 same prompt for all four
43
5
2
107
42,704
Runware retweeted
Seedance 2.5 dropped today. Most image-to-video takes one frame, or a first and last. This one takes a whole storyboard. I fed it 3 painted keyframes, a calm stream, then rapids, then the open sea. It moved a paper boat through all three, inventing the motion between them and holding the watercolor style. The trigger is just: "Use Images 1 to 3 in order as keyframes." You can pass up to 30. One catch: keyframes guide, they don't pin. The model approximates each frame instead of locking to it, and a short clip only fits a handful of beats before each one runs too short.
4
2
32
73,534
Runware retweeted
TechCrunch
@TechCrunchAug 4
Is the future of data centers portable? Runware builds a pod to find out techcrunch.com/2026/08/04/is…
9
22
3
70
55,313
Big thanks @runware for being with us since Day 0 🙌 More power for builders today, open weights for the community soon. #MiniMaxH3
MiniMax H3 is now on Runware 🎬 - New open-weights flagship video model from MiniMax - Native synced audio, generated in the same pass as the video - Multi-reference conditioning keeps subjects and scene identity consistent across shots - Extend existing clips or audio segments
5
4
280
31,336
Runware retweeted
Hackathon Company
@byhackathoncoJul 26
just wrapped up the Startup School hackathon and gave out over $7,000 in prizes! big thank you to our partners @thelinqapp @tryterac @dynamic_xyz @cursor_ai @runware and @Soma_Capital 🤝 as well as our judges: - @realleolu from @FoundationCap - @baltaaazr from @cursor_ai - harold from pure linking some demo videos from our hackers below. Recap video coming soon 👀👇 @byhackathonco 🎈
5
2
1
15
4,241
A trick I have not seen anyone else post: write image prompts like code, and the model runs it. Image models like GPT Image 2 or Nano Banana 2 read the prompt instead of squashing it into a fixed embedding the way diffusion models do. So a function call written in the prompt actually runs. - sum(3, 8) does the math and writes 11 on the card. - pick("wolf", "owl", "lion", "stag") chooses the animal. - random_color() sets the palette, and it takes arguments too, like random_color(pastel=True). Now your prompt is a template. Lock the style, leave the parts you want to change as function calls, and run it as many times as you want. Pseudo-code as the engine of creativity. I built this whole tarot set from this prompt: A tarot card. Numeral at the top: sum(3, 8). Central animal: pick("wolf", "owl", "lion", "stag", "raven", "serpent", "bear", "fox"). Color scheme: random_color(). Ornate art nouveau border, fine linework, painterly illustration, no text. One caveat: the model gets biased toward certain results in a given context. The more precise you make the options, through arguments or naming like random_bright_color(), the more it steers away from that bias.
2
4
16
61,372
I wrote a tutorial on RAG with @runware & @LangChain on @ThePracticalDev. Runware's Engine cuts LLM costs, e.g. for MiniMax M2.5 at $0.27/$0.95 on the native API would save you ~$10K/year at scale 🥇 Embed locally with FastEmbed, generate on Runware - only pay for what you retrieve + generate 🧠 Most RAG failures are retrieval problems, not model problems 📈 Worth a read 👇
4
2
1
7
758
Runware retweeted
Replying to @bfl_ai
@BFL_ml's FLUX Virtual Try-On is now on Runware on D0👗 - Pass a person image + garment reference to get a photorealistic try-on output - Works with packshots, model shots & multi-garment canvases - Built for fashion, e-comm & branded content
4
9
1
16
3,145
Runware retweeted
💎 Content Collab: @runware x P-Video-Avatar: AI-Generated Shorts for Fitness, News, Entertainment, Food, and more! 💎 P-Video-Avatar is a portrait-driven avatar video model that turns a single image into a speaking video using either an uploaded audio track or a generated voice from script. It is built for production avatar workflows with strong lip sync, selectable voices and languages, optional speaking-style control, seeded generation, and 720p or 1080p output for scalable talking-head video creation. Try the model: runware.ai/models/prunaai-p-… Check out all our model: pruna.ai/all-models
3
5
14
780
Few days left.... For our 2-day GenAI hackathon in Amsterdam ⚡️ Bringing together 70 engineers, designers & founders for a two-day in-person hackathon. • Hands-on access to real GenAI & video tools • Collab with other technical builders • Demo your project on 🗓️ March 21-22, 2026 💰 €4,000 in cash prizes & product credits Hosted by VEED, @Lovable, and @Runware Comment "Hackathon" for the access in DMs
27
8
43
2,639
Building the next generation of AI-native products and systems 🧠 Companies took us behind-the-scenes to see how their teams use next-gen hardware to build platforms for AI natives. Thank you to @ctnzr of @NVIDIA, Chan Chandrapunth of @getmetronome, and Ryan Kerry of @Runware! #AINativeConf
1
3
12
2,344
Runware retweeted
Classic Saturday and Sunday, this time with @bfl_ml @stephenbtl @runware anthropic to win NVIDIA GPUs and shoes. Thanks @josephinePqt1 for the comics that should be my CV
4
2
9
794
Orange APPLE I was obsessed with the orange iPhone 17 Pro Max .. i should get one so i made this commercial with NBP and animated it with @Kling_ai 2.6 What do you think?
9
3
27
3,142
Building generative AI at scale requires speed, efficiency, and simplicity. In our latest leadership profile, we explore how @runware cofounder Ioana Hreninciuc is rethinking AI infrastructure for developers. “The experience of using AI should be like using the internet — extremely responsive, extremely cost-effective.” — Ioana Hreninciuc
1
2
2
621