Meta Launches Spark 1.1 Agentic Coding Model
Curated by the Inblix editorial team
Meta just released Spark 1.1, a multimodal AI designed for agentic coding tasks. It can handle complex reasoning, manage digital workflows, and push new features into enterprise systems. While Meta is a bit late to the game — Anthropic and OpenAI have had similar models for a while — this launch is still a big deal. The pricing is competitive: $1.25 per million input tokens and $4.25 per million output tokens, roughly on par with rivals like Claude and GPT. Meta is pitching the model at enterprises looking to automate coding, fix bugs, and run large code migrations. CEO Mark Zuckerberg even returned to X after three years to hype it up, calling it “a strong agentic and coding model at a very low price.”
Why it matters: With multiple AI models dropping this week alone, Meta’s entry further intensifies the affordability race, forcing rivals to compete on both capability and cost — a win for businesses looking to deploy AI at scale.
💡 Key Takeaways
- Meta's Spark 1.1 focuses on agentic coding, automating complex workflows and enterprise tasks.
- Pricing is aggressive at $1.25 per million input tokens and $4.25 per million output tokens, matching competitors.
- CEO Mark Zuckerberg broke a three-year social media silence to promote the model, signaling its strategic importance.
Keep reading: See related articles below for more coverage on this topic.
Get smarter about AI
The sharpest AI news, curated daily. Delivered free to your inbox.