Axios is reporting that Reflection AI is about to release an extremely capable open weight model expected to compete with the top Chinese open weight ones. Reflection have been paying Elon $150 million a month for compute at Colossus since July, on top of a $1 billion compute deal with Nebius, so a strong model is certainly possible. Sources also told Axios that multiple open source models from other unnamed American labs are also planned to release this month. American OSS Renaissance about to begin? Howard Lutnick has been hoping for something like this for some time, and at one point was reportedly even considering direct government funding to get the ball rolling. The USG wants American open models to compete with Chinese models globally.Read original on X
StoryFirst seen
Reflection AI nears open-weight model release, Axios reports
The startup has paid Elon Musk's Colossus $150 million a month for compute since July, on top of a $1 billion Nebius deal.
Everything it touches
3 names 2 voices 2 source posts
The story’s neighborhood
Explore the graph, or choose a connection below.
Follow the evidence
The source posts
This brief comes from Tech Twitter’s edited headlines. The posts below are the sources grouped with the story; statements in them belong to their authors.
Would be a real shift - Chinese labs make up 90%+ of open-weight tokens on OpenRouter. Here’s to American open-weight!Read original on X
How the headline evolved
Reflection AI nears open-weight model release, Axios reports
Reflection AI nears open-weight model release, Axios reports
Reflection AI nears open-weight model release, Axios reports
Reflection AI nears open-weight model release, Axios reports
Reflection AI nears open-weight model release, Axios reports
Reflection AI nears release of open weight model, Axios reports
Reflection AI nears release of open weight model, Axios reports
Reflection AI nears open weight model release, Axios reports
Shared connections
Related stories
DeepSeek releases faster, cheaper update to its Flash model
The update is already live across major inference providers, keeping price pressure on in the open-model race.
Meta ships Muse Wonder frontier model amid rocky rollout
Early users point to strength on personal tasks and long-term memory while the team publicly fixes inference bugs, showing frontier launches now land before the software is stable
Four Gemini agents beat Minecraft with no human help
The agents wrote thousands of programs, coordinated over chat, and debugged each other's errors across a 27-hour run, offering a concrete look at spontaneous agent teamwork.
Google's Gemini 4 Argon tops benchmarks and beats Minecraft
Benchmark records, fast third-party availability, and autonomous agents that coordinated without any human help put Google's newest model at the center of the day's AI conversation