smolagents now support vLLM! As one of the most popular local inference solutions, the community had been asking us to integrate vLLM: after a heavy refactoring of our LLM classes, we've just released smolagents 1.11.0, with a brand new VLLMModel class. Go try it and tell us
LLMS
-

Multi-Agent Example with Local Models by RLanceMartin
By
–
Great stuff from @RLanceMartin getting a multi agent example with local models
-
LLMs: Free Access to Humanity’s Best Knowledge Anytime
By
–
I feel like people don’t appreciate enough that LLMs actively harness the best of what humanity knows to help you be your best self, any time of day for free. We’ve never had a technology that does that. And it’s the exact antidote to the tendency to promote extremes that the
-

Google Gemini impresses and threatens startups, making a strong comeback
By
–
It's far from perfect yet, but go ahead, @Google, they are too hot. The number of startups they will auto-kill headshot if they continue on this path with Gemini, it's going to be insane. Exactly what I expect from a giant like Google making a comeback.
-
How to test Gemini Flash 2.0 image generation
By
–
Now go to https://
aistudio.google.com and select model "Gemini Flash 2.0 Image generation" And test it out. -

How to Use Gemini API to Generate Text and Images
By
–
5/ Developers Can Try It NOW Google just opened experimental access via Google AI Studio & the Gemini API. You can start building today. Here’s a simple code snippet to generate text + images in one go:
-
Gemini 2.0 Flash improves text rendering in AI-generated images
By
–
4/ Better Text Rendering in Images
— God of Prompt (@godofprompt) 15 mars 2025
AI text in images usually looks like gibberish. But Gemini 2.0 Flash is much better at rendering long, legible text.
This means ads, posters, and social media graphics now look professional. pic.twitter.com/UOJldIcydk4/ Better Text Rendering in Images AI text in images usually looks like gibberish. But Gemini 2.0 Flash is much better at rendering long, legible text. This means ads, posters, and social media graphics now look professional.
-
Conversational Image Editing using Gemini 2.0 Flash
By
–
2/ Conversational Image Editing
— God of Prompt (@godofprompt) 15 mars 2025
No more guessing the perfect prompt. With Gemini 2.0 Flash, you can edit images step by step in natural conversation.
“Make the background brighter.” → “Add a sunset.” → “Make the cat fluffier.” ✅ pic.twitter.com/X1qs2hHooM2/ Conversational Image Editing No more guessing the perfect prompt. With Gemini 2.0 Flash, you can edit images step by step in natural conversation. “Make the background brighter.” → “Add a sunset.” → “Make the cat fluffier.”
-
Gemini 2.0 Flash enables consistent AI-generated storybook visuals
By
–
1/ Text + Images in One Model
— God of Prompt (@godofprompt) 15 mars 2025
You can now prompt Gemini 2.0 Flash to generate stories with matching illustrations and it keeps characters & settings consistent.
Imagine writing a children’s book where AI handles both the text and the visuals. pic.twitter.com/WfMJjC2yh61/ Text + Images in One Model You can now prompt Gemini 2.0 Flash to generate stories with matching illustrations and it keeps characters & settings consistent. Imagine writing a children’s book where AI handles both the text and the visuals.
-

Google Releases Gemini 2.0 Flash with Native Image Generation
By
–
This is WILD. Google just dropped Gemini 2.0 Flash, and it can now generate images natively. Here’s what that means for the future of AI:
