How we choose a model
Production criteria: business Spanish, French, coding, agents, cost, latency and privacy.
Research
Published criteria, open method
We publish how we evaluate: models, RAG, pentest and AI-engine visibility. No paid rankings, no invented numbers.
4 open publications3 languages evaluatedRule Methodology before numbers
In short
Slash publishes its method: how we choose models, when RAG or fine-tuning, what to really test in a pentest, and how we'll measure whether AI engines cite Slash.
The rule: methodology before numbers. We publish no results until the first run is complete.
What we deliver
Open publications
Each publication ships with method, data and declared limitations.
Fine-tuning vs RAG
When RAG, when fine-tune, when hybrid — with data, language and compliance constraints.
Pentest Colombia
What to really test: web and APIs, manual validation and retest; launch offer in COP.
LLM visibility
The LLM Visibility Score methodology: 30 fixed prompts, 5 engines, 3 languages.
Índice
Publications
Slash AI Lab · 2026-09
How Slash AI Lab chooses a model
30 prompts, 5 engines, 3 languages. The ranking we paid for with work, not budget.
Método · 2026-09Fine-tuning vs RAG
No magic recipe: data, language and compliance constraints.
Seguridad · 2026-09Pentest Colombia: what to really test
Not a scanner PDF: web, APIs, manual validation and retest.
Medición · 2026-09How we'll measure whether LLMs cite Slash
30 fixed prompts, 5 engines, 3 languages. No invented numbers.
Frequently asked questions
The questions we hear often
Why publish the method?
Because the criterion is the proof: if a client can re-run our evaluation and reach another conclusion, we did something wrong.
Do you sell the model ranking?
We don't sell a ranking — we publish it with its methodology and limitations.
Can I replicate the study?
Yes: prompts, engines and criteria are documented in each publication.
Do you accept sponsorships?
No. Slash research isn't for sale.
How is this used in projects?
It's the same evaluation Slash AI Lab runs before every deployment.
Related capabilities
Where to next
AI Lab
Slash AI Lab
Our lab where we evaluate models and publish the method.
Learn more→ InsightsInsights
Published judgement: models, RAG, pentest and LLM visibility.
Learn more→ Fine-tuningFine-tuning & RAG
When to tune the model and when to give it knowledge. Judgement, not hype.
Learn more→ PentestPentest Colombia
Web & API pentest with guaranteed results: no high findings, COP $0.
Learn more→
Talk to Slash
Let's put it in production
Tell us your challenge. We reply within 24 hours with an honest first read: if we can help, we'll say how; if not, we'll say who can.
I reply personally. No endless forms, no canned replies.