ModelsHaiku 5.5: Cheaper AI can lose the evidence your strongest model needs
Anthropic brings adjustable reasoning to its small-model tier. The savings depend on what survives the handoff to a larger model, not just the price of the first answer.
Models and Architectures
Model releases and architectures: what changes in capability, efficiency and control.
LLMs · VLMs · VLAs · World models · Image, video & audio · Embeddings
8 articles
Models
ModelsMistral Large 4: Smarter reasoning cannot fix a misread document
The multimodal preview highlights a costly blind spot in document AI: a plausible answer may come from the wrong chart, label or footnote. Test what the model saw before scoring what it concluded.
ModelsReflection’s Beam: Shorter reasoning only saves money if the answer still works
Reflection trains its model to avoid unnecessary reasoning tokens. The business test is whether it removes wasted work without increasing failed attempts or human correction.
ModelsEmbeddingGemma 2: Better AI answers can start with finding the missing evidence
Google’s compact model searches across text, images, video and audio in a shared representation. Its useful advantage may be a simpler retrieval pipeline, provided it finds the records that actually matter.
ModelsDeepSeek V4.1 Flash: Cheaper AI memory could change what your model remembers
DeepSeek separates the work of reading from answering and reports a smaller context cache. For long-running systems, the opportunity is to retain useful evidence at a lower cost—not simply generate tokens faster.
ModelsGPT-6.1 Sol: The biggest AI saving may come from changing your default model
OpenAI positions Sol near its pricier Astra on selected evaluations. The practical opportunity is moving routine work to a cheaper model while identifying the exceptions that still need more capability.
ModelsGemini Robotics 2: A new robot should not mean starting AI training again
Google DeepMind’s transfer claims point to a potentially scalable model business: reuse learned skills across different bodies. The test is total adaptation effort, including data collection and validation.
ModelsPhysical Intelligence’s π0.7: Teach a robot a new task before retraining it
The steerable model explores whether richer instructions can reuse skills already learned. The business opportunity is fewer new demonstrations; the unresolved cost is the human guidance needed to make it work.
No articles in this topic yet
Explore the other topics in this desk.