Big day for our team at Meta. Today, we launched an improved Muse Spark model with major improvements in agentic workflows, coding, computer use, and multimodal perception.
Even better? It's now available to developers through Meta Model API.
👉https://t.co/FBM9FxPbnK 🚀
Developer choice is core to what we're building.
We’re excited to share that Muse Spark 1.1 is now available on @OpenRouter for US-based developers.
Get started today 👉 https://t.co/DTyYKrTgNK
@OpenRouter availability was the #1 thing devs asked for after we shipped Muse Spark 1.1 last week.
Excited to get it in more hands and see what the community builds
Developer choice is core to what we're building.
We’re excited to share that Muse Spark 1.1 is now available on @OpenRouter for US-based developers.
Get started today 👉 https://t.co/DTyYKrTgNK
Muse Spark 1.1 from @Meta is live in OpenClaw.
Update to OpenClaw v2026.7.1 and experience Meta's new multimodal reasoning model for agentic coding, tool use, and computer-use workflows.
OpenClaw v2026.7.1 is out, with 3,063 contributions from 532 contributors!
• Web UI and onboarding overhauls
• Major iOS, Android and MacOS app work
• Telegram, Slack, Discord, and iMessage improvements
• GPT-5.6, Muse Spark 1.1, and MORE
https://t.co/NAOyTZb4gB
OpenClaw v2026.7.1 is out, with 3,063 contributions from 532 contributors!
• Web UI and onboarding overhauls
• Major iOS, Android and MacOS app work
• Telegram, Slack, Discord, and iMessage improvements
• GPT-5.6, Muse Spark 1.1, and MORE
https://t.co/NAOyTZb4gB
Got inspired by @Clad3815 to see how Muse Spark 1.1 performs as a vision-only computer use agent: playing Pokémon Emerald from raw screenshots. Same constraints as his GPT-5.6 SOL run: no RAM reads, no map data, no walkthrough guides. The only knowledge that persists is what the player agent perceives and writes into its own atlas to keep track of its navigation, questlog, and NPCs.
Harness is on v16 - every version fixed something agent audits caught in the logs watching it play. Built on the Meta Model API CUA & perception-grounding docs: look → reason → act → look, with pixel-precise grounding and the Responses API carrying reasoning between moves.
This is definitely not an optimal build/harness, but going to see how far this thing goes before making more improvements
AI developers - we’ve published a technical guide covering how to get started with Muse Spark on Meta Model API. Muse Spark is a multimodal reasoning model built for agentic tasks, coding, computer use and long-context workflows.
See what you can build 👉 https://t.co/d3raFYsyxp
Big day for our team at Meta. Today, we launched an improved Muse Spark model with major improvements in agentic workflows, coding, computer use, and multimodal perception.
Even better? It's now available to developers through Meta Model API.
👉https://t.co/FBM9FxPbnK 🚀
Meta just released Muse Spark 1.1 and is the new SOTA on MedScribe and TaxEval, taking the top spot from Fable 5 while being 10x cheaper and twice as fast. Meta currently holds the top 2 spots on TaxEval
It is also the new #1 on Harvey's Legal Agent Bench, dethroning Grok 4.5 less than 24 hours after it took the top spot.