Microsoft's MAI-Code-1-Flash, a lightweight coding model for GitHub Copilot in VS Code, outperforms Claude Haiku 4.5 and GPT-5.4 Mini on code quality while using fewer tokens, according to aggregate d…
GitHub reported degraded availability for multiple GPT models in Copilot products and IDE surfaces on July 25, including GPT-5.2, GPT-5.3-Codex, GPT-5.4, GPT-5.4 Mini, GPT-5.6 Sol, GPT-5.6 Terra, and …
Microsoft AI on July 23rd, 2026 announced the rollout of two new MAI models built with a hill-climbing training pipeline, integrated into GitHub Copilot and Microsoft Excel. The models claim to match …
Seven models ran the dynaconf__dynaconf-1225 SWE-bench-Live task with both answer doors shut; three solved it and four failed, with costs ranging from $0.78 (gpt-5.4-mini) to $47.18 (opus-4.8). The co…
At the AI Engineer World's Fair, workshops shifted from RAG and prompt engineering toward evals and open models, reflecting a broader industry focus on measuring and trusting AI outputs. A developer a…
A new architecture using a nano-model router that dispatches knowledge-worker tasks to either a frontier model or a small, cheap model achieves near-frontier quality at a fraction of the cost, ranking…
OpenAI released GPT-5.4 Mini at $0.75 per million input tokens and $4.50 per million output tokens, while GPT-5.4 costs $2.50 and $15, a 3.3x price multiplier. A developer's task-by-task analysis show…
A new preprint from Cornell University researchers reveals that chatbots from OpenAI, Anthropic, and Google repeatedly generate stories featuring the character "Elias Thorne," a lighthouse keeper, due…
Ten strategies to reduce costs associated with LLM API usage, emphasizing that users pay for input and output tokens rather than the inference itself. Key recommendations include selecting appropriate…