Some LLM takeaways for 2025: reasoning as a signature feature, coding agents were useful, subscriptions hit $200/month, and Chinese open-weight models impressed
It's that time. It's been a hell of a year. — At the start we barely had reasoning models. X: Simon Willison / @simonw : Here's my enormous round-up of everything we learned about LLMs in 2025 - th...
An overview of AI in 2025, including arguments for and against above-trend model capabilities growth, the state of evals, and the safety of reasoning models
Gavin Leech / LessWrong : X: @g_leech_ , @g_leech_ , and @patrick_oshag X: Gavin Leech / @g_leech_ : My summary of the year in AI [image] Gavin Leech / @g_leech_ : ADeLe really is an amazing eval [im...
An analysis of 100T+ tokens from the past year shows reasoning models now represent over half of all usage, open-weight model use has grown steadily, and more
this is not a model I hear much about. [image] @openrouterai : We collaborated with @a16z to publish the **State of AI** - an empirical report on how LLMs have been used on OpenRouter. After analyzing...
An analysis of 100T+ tokens from the past year shows reasoning models now represent over half of all usage, open-weight model use has grown steadily, and more
An Empirical 100 Trillion Token Study with OpenRouter — Malika Aubakirova*Alex Atallah†Chris Clark†Justin Summerville†Anjney Midha*
Gemini co-lead Oriol Vinyals says Gemini 3's gains come from better pre-training and post-training, contradicting the idea that pre-training gains are falling
which we discussed in our NeurIPS '25 talk with @ilyasut and @quocleix—the team delivered a drastic jump. The delta between 2.5 and 3.0 is [image] Andrej Karpathy / @karpathy : I played with Gemini 3 ...
OpenAI releases gpt-oss-safeguard, its open-weight reasoning models for safety classification tasks, available in 120B and 20B parameters, under Apache 2.0
New open safety reasoning models (120b and 20b) that support custom safety policies. — Today, we're releasing a research preview …
Mira Murati's Thinking Machines Lab launches its first product, Tinker, an API for fine-tuning language models, in private beta, with support for Qwen and Llama
Today, we are launching Tinker, a flexible API for fine-tuning language models. Moneycontrol : Ex-OpenAI CEO Mira Murati stealth AI lab launches its first ever product Matthias Bastian / The Decoder :...
xAI launches Grok 4 Fast, a multimodal model with a 2M context window and a unified architecture that combines reasoning and non-reasoning modes
Pushing the Frontier of Cost-Efficient Intelligence — We're thrilled to present Grok 4 Fast, our latest advancement in cost-efficient reasoning models.
The price per token for AI models has fallen, but costs for developers are rising as newer reasoning models require more tokens to complete tasks
With models doing more ‘thinking,’ the small companies that buy AI from the giants to create apps and services are feeling the pinch X: @ericjhonsa , @emollick , @mims , @mims , @mims , @mims , and @m...
Sam Altman says OpenAI will bring back GPT-4o to ChatGPT and raising reasoning model rate limits for free and Plus users, as usage of reasoning models increases
The move is a stunning reversal, proving that even the most powerful AI company can't ignore a mutiny from its loyal user base.