Models & Labs
GPT-5.6 Triples Scores on ARC-AGI-3 Benchmark
OpenAIhigh confidence
Why it matters
- →Demonstrates the impact of configuration adjustments on AI performance.
- →Highlights the potential for efficiency gains without major architectural changes.
- →Emphasizes the importance of optimizing existing AI capabilities.
OpenAI has reported a substantial improvement in the performance of its GPT-5.6 model on the ARC-AGI-3 benchmark. By enabling two specific API settings, the model's scores have tripled, showcasing enhanced reasoning and efficiency. This advancement demonstrates how strategic adjustments in configuration can significantly boost AI performance without altering the model's core architecture. The results emphasize the potential for optimizing existing AI systems to achieve superior outcomes.
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 Release
Matt Wolfe · April 24, 2026 · Same story
- 3OpenAI Launches GPT-5.5 as Advanced AI Model
AI News · April 29, 2026 · Same story
- 4OpenAI launches GPT-5.6 amid regulatory scrutiny
The Verge AI · June 26, 2026 · Same story
- 5OpenAI Unveils GPT-5.6 with Enhanced Capabilities
OpenAI · July 9, 2026 · Same story
- 6OpenAI Launches GPT-5.6 with New Features
The Rundown AI · July 10, 2026 · Same story
- 7OpenAI Unveils GPT-5.6 with Sol Model
Fireship · July 10, 2026 · Same story
- 8GPT-5.6 Enhances AI Efficiency and Intelligence
OpenAI · July 29, 2026 · Same story
What happened next
- 9OpenAI Launches GPT-6 Astra, Claims AGI Milestone
WIRED AI · September 3, 2026 · Same story
- 10OpenAI unveils GPT-6 Astra, claims AGI milestone
The Rundown AI · September 4, 2026 · Same story
- 11OpenAI Introduces GPT-6 with Upgraded Computer Use
Matt Wolfe · September 24, 2026 · Related
- 12OpenAI releases GPT-6.1 Sol, cuts costs while boosting safety
TechCrunch AI · September 29, 2026 · Related
- 13OpenAI Releases GPT-6.1 Terra for Computer Use
Duncan Rogoff · October 26, 2026 · Related
More in Models & Labs
Models & Labsother
llama.cpp b12340 adds CUDA 13 and ROCm 10.0
This release quietly closes the hardware gap for local inference by adding default support for CUDA 13 and ROCm 10.0 across Linux and Windows. Users with newer NVIDIA GPUs or AMD accelerators no longer need to compile from source to get performance, while Apple Silicon builds now ship KleidiAI disabled by default, shifting the focus back to standard ARM optimizations. The inclusion of Snapdragon support for Linux and Android marks a significant step toward on-device AI beyond the x86/ARM desktop ecosystem. It is less about new features and more about ensuring llama.cpp remains the universal runtime for every available accelerator.
llama.cpp Releases
Models & Labscoding
llama.cpp b12342 adds few-row MMQ for speculative decoding
Speculative decoding just got a significant speed boost on NVIDIA GPUs. This release introduces a specialized kernel for 'few-row' matrix multiplications, targeting the small batch sizes typical of verification passes in speculative decoding workflows. By optimizing these specific operations for quantized formats like Q8_0 and IQ4 variants, llama.cpp reduces latency where it matters most during token generation. The change is CUDA-only for now but leverages hardware capabilities from Ampere onwards, making local inference faster without requiring new model architectures or complex configuration changes.
llama.cpp Releases
© TechCrunch AIModels & Labsmodels
Google deploys Gemini 4 Hydrogen for faster Search
Google is quietly shifting its search infrastructure by replacing older fast models with Gemini 4 Hydrogen, a tier designed specifically for speed and cost-efficiency rather than raw reasoning power. This model completes tasks 2.5 times faster than its predecessor while beating the more powerful Argon on specific intelligence benchmarks, proving that specialized efficiency can outperform brute force in high-volume workflows. By integrating this into Search's fan-out technique, Google now breaks down complex queries into subqueries with greater precision and speed, delivering richer visual responses without the latency of heavier models. This marks a strategic pivot toward agentic execution layers within consumer search, making AI Mode feel snappier for everyday multi-part questions.
TechCrunch AI