Independent AI Observatory

Artificial intelligence, with context.

Explore models, discover tools and understand the advances that are changing technology.

Explore: Border models Tools Local AI
Animated robot workshop. Use the buttons Give energy, Achís! and Create an idea.
A spark is about to become something.
01 / NEWS

What matters, explained

See all the news ↗
02 / EXPLORE

Start with your question

Six routes from headline to informed decision.

M
CATALOG

Models

Verified profiles, capabilities and versions.

↗ T
DIRECTORY

Tools

What they solve and for whom.

↗ A
SYSTEMS

Agents

Planning, memory and action.

↗ D
PRACTICE

Development

APIs, patterns and evaluation.

↗ L
ON YOUR TEAM

Local AI

Privacy and control without cloud.

↗ R
EVIDENCE

Investigation

The papers, without losing rigor.

↗
03 / MODELS

A comparable view

Select up to three models to contrast their orientation and modalities.

OVERIFIED · 16 SEP 2026
OPENAI

GPT‑6 Astra

Generalist frontier model oriented towards complex tasks, use of tools and highly demanding professional work.

TextImage
Profile
Generalist frontier
State
Available
View profile →
AVERIFIED · 16 SEP 2026
ANTHROPIC

Claude Opus 5

Model oriented to code and knowledge work, with controls that allow effort to be adjusted to each task.

TextImage
Profile
Knowledge and code
State
Asset
View profile →
GVERIFIED · 16 SEP 2026
GOOGLE

Gemini 3.8 Flash

Flash model for long-running software engineering, autonomous agents, and complex business flows.

TextImageVideoAudio
Profile
Fast multimodal
State
Stable GA
View profile →

i Profiles show their verification date and always link to the primary source. Missing data is never inferred.

FEATURED GUIDE

How to choose a local model without blindly buying hardware

Memory, quantization, context and speed: a practical route to fit the model to the computer and not the other way around.

01 Defines the use02 Measure memory03 Evaluate quality
Read full guide ↗
RESEARCH EXPLAINEDRecent readings
EDITORIAL PIPELINE

Why more context doesn't always produce better answers

Attention, useful signal and practical limits.

→
EDITORIAL PIPELINE

Agent evaluation needs to measure processes, not just results

A guide to interpreting long benchmarks.

→
EDITORIAL PIPELINE

Small and multimodal models: the race moves to the device

What changes when the inference is local.

→
04 / TOOLS

Selection with criteria

Less infinite directory; more context about the problem each tool solves.

C
DEVELOPMENT

Code Wizards

To understand, review and maintain repositories.

Collection in preparation
R
INVESTIGATION

Search with sources

To synthesize evidence without losing traceability.

Collection in preparation
V
CREATION

Image and video

For ideation, production and visual editing.

Collection in preparation
05 / EVOLUTION

From chat to systems that act

An editorial reading of how the product focus has changed. It is not a benchmark: it summarizes recurring capabilities in the official announcements analyzed.

Consolidated capacityEmerging focus
06 / USE CASES

Decide from the problem

Each route connects a need with evaluation criteria, not with a specific brand.

01PRODUCT TEAMS

Analyze feedback without losing traceability

Privacy, long context and citations as minimum criteria.

View relevant criteria →
02DEVELOPMENT

Maintain a large repository

Understanding the code, tools and automatic verification.

Compare capabilities →
03SENSITIVE ENVIRONMENTS

Process documents without sending them to the cloud

Local memory, quantization and control of the environment.

Open local AI guide →