Models & Labs
OpenAI's GPT-5.6 Sol Model Released to Limited Partners
The Rundown AIhigh confidence
Why it matters
- →GPT-5.6 Sol represents a significant leap in AI reasoning capabilities.
- →Government-controlled access could set a precedent for future AI releases.
- →Limited availability may drive demand for alternative AI solutions.
OpenAI has launched GPT-5.6 Sol, its most powerful AI model to date, but access is currently limited to around 20 vetted partners due to U.S. government restrictions. The model features advanced reasoning capabilities and an 'ultra' mode for handling complex tasks. While it outperforms previous models in certain benchmarks, its restricted release highlights a trend towards government-controlled access to cutting-edge AI technologies. This could lead to delays in broader availability, prompting companies and governments to seek alternative solutions.
Read originalThe story around this
Earlier coverage that leads up to this article, and what followed. Lines connect each piece to the closest one after it, converging here.
How we got here
- 1OpenAI's GPT-5.5 Testing Shows Promising Results
Wes Roth · April 21, 2026 · Same story
- 2OpenAI announces GPT-5.5 availability
Matthew Berman · April 23, 2026 · Same story
- 3OpenAI Announces GPT-5.5 Release
Matt Wolfe · April 24, 2026 · Same story
- 4OpenAI Releases GPT-5.5 and ChatGPT Images 2.0
The AI Advantage · April 24, 2026 · Related
- 5OpenAI Launches GPT-5.5 as Advanced AI Model
AI News · April 29, 2026 · Same story
- 6OpenAI to Launch GPT-5.5-Cyber for Select Users
The Verge AI · April 30, 2026 · Related
- 7OpenAI Enhances Voice Agents with Real-Time Models
The Rundown AI · May 8, 2026 · Related
- 8OpenAI Unveils GPT-5.5 Instant for ChatGPT
The AI Advantage · May 8, 2026 · Related
What happened next
- 9OpenAI Launches GPT-5.6 with New Features
The Rundown AI · July 10, 2026 · Same story
- 10OpenAI Unveils GPT-5.6 with Sol Model
Fireship · July 10, 2026 · Same story
- 11GPT-5.6 Enhances AI Efficiency and Intelligence
OpenAI · July 29, 2026 · Same story
- 12OpenAI Unveils GPT-6 Astra Model
OpenAI · September 3, 2026 · Same story
- 13OpenAI Launches GPT-6 Astra, Claims AGI Milestone
WIRED AI · September 3, 2026 · Same story
Follow this story
Open the full story →OpenAI delays GPT-5.6 after government request
19 developments
- Jun 25 · The Verge AI
OpenAI delays GPT-5.6 after government request - Jun 25 · TechCrunch AI
White House Influences OpenAI's GPT 5.6 Release Strategy - Jun 26 · The Rundown AI
White House Limits OpenAI's GPT-5.6 Release - Jun 26 · OpenAI
OpenAI Previews GPT-5.6 Sol Model↳ OpenAI previews GPT-5.6 Sol, highlighting enhanced coding, science, and cybersecurity capabilities alongside an advanced safety stack. - Jun 26 · TechCrunch AI
U.S. Government Tightens Control on AI Model Releases↳ The government approval process may stall data center expansions and impact AI lab revenues. - Jun 26 · The Verge AI
OpenAI launches GPT-5.6 amid regulatory scrutiny - Jun 26 · WIRED AI
OpenAI Delays GPT-5.6 Release Due to US Government Request - Jun 26 · Matt Wolfe
US Government Involvement in OpenAI Model Releases↳ THE US GOVERNMENT IS NOW INVOLVED IN OVERSEEING OPENAI'S MODEL RELEASES. - Jun 26 · TechCrunch AI
OpenAI Limits GPT-5.6 Rollout After U.S. Request - Jun 26 · Lev Selector
US Government Delays OpenAI GPT 5.6 Release - Jun 29 · The Rundown AI
OpenAI's GPT-5.6 Sol Model Released to Limited Partners (This article) - Jun 30 · The AI Daily Brief
OpenAI Previews GPT-5.6 to Trusted Partners↳ OpenAI previews GPT-5.6 to trusted partners, sparking discussions on benchmarks, safety, and government review. - Jul 2 · AI Explained
OpenAI Releases GPT 5.6 Sol↳ The new article adds nothing not already covered by the June 29 beat stating GPT-5.6 Sol was released to limited partners. - Jul 3 · Matt Wolfe
OpenAI Previews GPT-5.6 Sol - Jul 3 · Lev Selector
OpenAI Releases GPT-5.6 Models: Sol, Terra, Luna - Jul 9 · OpenAI
OpenAI Unveils GPT-5.6 with Enhanced Capabilities - Jul 9 · GitHub Changelog
OpenAI's GPT-5.6 Models Now in GitHub Copilot↳ GPT-5.6 models Sol, Terra, and Luna are now integrated into GitHub Copilot for developers. - Jul 9 · The Verge AI
OpenAI launches GPT-5.6 and ChatGPT Work↳ OpenAI launches GPT-5.6 publicly with government approval and introduces ChatGPT Work agent to compete with Anthropic's Claude Cowork. - Jul 9 · Matthew Berman
GPT-5.6 Announced for Box Integration↳ GPT-5.6 will be integrated into Box to enhance user workflows and productivity.
More in Models & Labs
Models & Labsmodels
llama.cpp b12029 adds CUDA 13 and Snapdragon support
This release quietly cements llama.cpp as the universal inference runtime by finally supporting NVIDIA's latest CUDA 13 stack alongside AMD's ROCm 10.0. The inclusion of native builds for Linux arm64 with Snapdragon chips, covering CPU, Adreno GPU, and Hexagon NPU, signals a serious push into mobile edge deployment. While Apple Silicon KleidiAI is disabled in this specific build, the broader hardware coverage means developers no longer need to compile from source for most modern accelerators. It’s less about new features and more about removing friction for anyone running local models on bleeding-edge or diverse hardware.
llama.cpp Releases
© The Verge AIModels & Labsother
Apple Intelligence arrives in China via iOS 27.3
Apple is finally unlocking its AI suite for Chinese users with the iOS 27.3 beta, marking a significant regulatory milestone after months of approval delays. The rollout relies on Alibaba’s Qwen model rather than OpenAI, integrating features like notification summaries and visual intelligence directly into the OS. By sidestepping the new Siri in favor of established text and image tools, Apple is prioritizing immediate utility over the full generative experience. This move effectively brings Apple’s AI capabilities to a massive new market, albeit with localized constraints that distinguish it from the global release.
The Verge AI
© Matt WolfeModels & Labsmodels
Google Releases Gemini 4 Carbon for Cost-Efficient Inference
Google launched Gemini 4 Carbon, a new model variant aimed at providing high performance at a lower cost for inference tasks.
Matt Wolfe