MAI-Image-2.6 & MAI-Image-2.6-Flash are now live on AIHubMix.
Microsoft's image models for creative workflows:
🔹2.6: fine creative control
🔹Flash: low-latency production
🔹Multi-reference editing
🔹Dynamic aspect ratios
Microsoft's MAI-Image-2.6-Flash takes #3 on the Artificial Analysis Image Editing Leaderboard, a significant jump over the previous generation’s Flash variant and joining MAI-Image-2.6 on the Pareto frontier for quality vs price
MAI-Image-2.6-Flash is an optimized version of
Muse Spark 1.3 vs Gemini 3.8 Flash
Same prompt: "A teddy bear deflating as it runs"🧸
Side-by-side AI video generation via AIHubMix.
Comparing visual quality, character consistency, generation time & API cost.
Muse Spark 1.3: $0.09 | 137s
Gemini 3.8 Flash: $0.06 | 205s
We asked GPT-6-Astra to build an Odyssey-inspired 3D level, play through it, and record the result.
The result:
8 platforms and 3 moons
Returning cap attacks, coins, and enemies
Checkpoints and touch controls
Original music and sound effects
Built with @threejs, Rapier physics,
GPT-6 Astra is now live on AIHubMix.
OpenAI’s most capable model for complex reasoning, coding, computer use, research & document creation:
🔹1.05M-token context
🔹Up to 128K output
🔹Low-to-max reasoning
🔹Async tool calling
🔹Mid-turn steering
GPT-6 Astra is now available to all Pro, Enterprise, and Business Premium users in ChatGPT Work and Codex. It's also live in the API.
It might take a few days to roll out to our Plus and Business users. Thank you for your patience.
