Lead · NEWS · 4 min

Mistral-Inference v1.4.0 Adds Vision via Pixtral-12B

Mistral AI released mistral-inference v1.4.0, adding multimodal vision support to its open-source inference library through a new model called Pixtral. The update enables the Pixtral-12B-2409 model to process images alongside text, accessible via a command-line interface or Python API after a standard pip upgrade.

Mistral-Inference v1.4.0 Adds Vision via Pixtral-12B

Research

Papers, benchmarks, and what we'd build differently after reading them.

All Research →
Research · May 19 · 6 min

Cross-Component Interference Undermines LLM Agent Stacking

Research · May 18 · 5 min

LATTE Framework Uses Task Graphs to Coordinate LLM Teams

Research · May 18 · 6 min

NeuroAgent Automates Multimodal Neuroimaging Pipelines

Research · May 14 · 7 min

CIVeX Verifier Targets Causal Gaps in Tool-Using Agents

Tutorials

Code-first dispatches. Most run in under fifteen minutes.

All Tutorials →
Multimodal Retrieval Pipeline with LlamaIndex AgentMesh Trust Layer
Tutorial · May 18 · 11 min

Multimodal Retrieval Pipeline with LlamaIndex AgentMesh Trust Layer

Per-Generation LLM Cost Tracking with Langfuse v4 SDK
Tutorial · May 18 · 10 min

Per-Generation LLM Cost Tracking with Langfuse v4 SDK

Wiring OpenTelemetry Spans into a LangGraph Agent for Cost Attribution
Tutorial · May 18 · 8 min

Wiring OpenTelemetry Spans into a LangGraph Agent for Cost Attribution

Wiring OpenTelemetry Spans into a LightRAG Retrieval Pipeline
Tutorial · May 18 · 7 min

Wiring OpenTelemetry Spans into a LightRAG Retrieval Pipeline

Tools this week View all →
Plumb 0.4
May 13
Native eval traces, OTLP exporter
lattice/agents v1.2
May 12
Replay scrubber, branch comparison
skiff-trace 0.9
May 10
Lightweight OTel for tool-using agents
harness-kit 2026.5
May 09
Dockerised runners with cgroup limits