Loading activity…
@loganmarkewich
Head of Open Source @llama_index
AI
LlamaIndex expects considerably faster inference for LightOnOCR-3-1B, but its announcement gives no numerical results.
AI
The person who added them notes low Muse Wonder latency and says Gemini pricing will double later in 2026.
AI
A user singles out Slack and Teams, arguing that directly pasted AI text removes human conversation from work.
AI
The launch post says the model combines text transcription, document layout, image descriptions and chart extraction.
AI
Anthropic says it costs around 75% less to run on average than Haiku 4.5 and improves coding and computer use.
AI
LlamaIndex says users can switch models in one line while keeping the same request and markdown output.
AI
LlamaIndex claims the agents outperform Opus 5.5 and GPT-6 Sol while being 30% to 4x cheaper.
AI
A user says the goal is to run something like SmolLM and understand every layer. They describe the work as refreshing, though even reshape() has proved complicated.
AI
The announcement describes Jeff as a drop-in replacement for jev based on Knowledgator’s GliFormer, and says only LLM/decoder-based models ranked higher.
AI
Its developer describes it as a drop-in replacement, claiming the same latency as jev at lower cost, with a mild hit to accuracy.
