LLMs·
The Quiet Architecture Shift Behind Llama 4
Beyond the launch posts, DeepSeek V4 is reshaping how product managers approach contract review. We talked to the people actually using it in production.
Elena Brost3 min read
The archive
All articles
100 stories from the Inference Daily desk.
The Quiet Discipline Behind Reliable Qwen 3 Apps
The interesting story is not the demo. It is the second month, when structured outputs either earns its keep or gets quietly rolled back.
Daniel Okafor3 min read
LLMs·
Why Claude 4.5 Sonnet Feels Different, Even When the Numbers Don't
The interesting story is not the demo. It is the second month, when small-model orchestration either earns its keep or gets quietly rolled back.
Daniel Okafor3 min read
Mistral Loses a Key Researcher to HubSpot
Beyond the launch posts, Claude 4.5 Sonnet is reshaping how sales teams approach onboarding. We talked to the people actually using it in production.
Daniel Okafor3 min read
LLMs·
Google DeepMind Releases Qwen 3: What the Benchmarks Actually Show
Beyond the launch posts, Mistral Large 3 is reshaping how product managers approach contract review. We talked to the people actually using it in production.
Daniel Okafor3 min read
LLMs·
Scaling Laws After Llama 4: A Field Report
The interesting story is not the demo. It is the second month, when tool-first agents either earns its keep or gets quietly rolled back.
Jonas Halvorsen3 min read
small-model orchestration: The Pattern Behind the Best Agents This Quarter
The interesting story is not the demo. It is the second month, when long-context workflows either earns its keep or gets quietly rolled back.
Jonas Halvorsen3 min read
Why founders Are Suddenly Standardizing on Linear AI
Spotify is the latest in a string of teams treating small-model orchestration as the default, not the experiment. Here is what they got right — and what they are still figuring out.
Daniel Okafor3 min read
Inside the New Wave of Perplexity Built for engineering teams
Intercom is the latest in a string of teams treating RAG-as-a-service as the default, not the experiment. Here is what they got right — and what they are still figuring out.
Yuki Tanabe3 min read