Galtea
Barcelona platform for generating test cases, simulating users and evaluating AI systems in regulated industries
Quick Overview
| Company | Galtea Solutions S.L. |
|---|---|
| Category | LLM Observability and Evaluation Tools |
| Headquarters | Barcelona, Spain |
| Founded | Not stated |
| EU Presence | Yes - Spain (EU) |
| Data Location | Cloud, or self-hosted or private tenant on Enterprise |
| Open Source | No |
| GDPR Compliant | Yes |
| Self-Hosting | Enterprise - self-hosted or private tenant |
| Main Features | Synthetic test case generation, conversation simulation, LLM-as-a-judge evaluation, red teaming, production monitoring, custom metrics, Python SDK, REST API |
| Pricing | Free plan (EUR 0, 1,500 credits/month); Enterprise on request |
| Best For | AI teams in banks, insurers and public bodies that must document how they tested a model |
| Replaces | LangSmith, Arize |
Detailed Review
Galtea starts from specifications rather than traces. You describe what the AI system should and should not do, and the platform generates test cases, simulated multi-turn conversations and red-team attempts, then scores the answers with evaluators.
The about page says the company emerged from research at the Barcelona Supercomputing Center. The terms are issued by Galtea Solutions S.L., a Spanish company registered in Barcelona, and the site lists ISO 27001 certification and GDPR data processing agreements.
Test generation and simulation
Thousands of test cases can be generated from requirements, policies and domain documents, so a dataset is not needed up front. Simulated users run realistic multi-turn journeys and adversarial behaviour before release.
Credit costs are published: red teaming and scenario test cases cost 1 credit, a judge evaluation 2 credits, a simulated conversation turn 3 credits and a gold-standard test case 5 credits.
Monitoring and audit trail
Quality, safety, compliance, cost and performance are tracked across versions and live interactions. The site stresses a traceable record for auditors, which fits the regulated-industry focus.
Plans
The free plan gives 1,500 monthly credits, one user, two products and 90 days of retention. Enterprise adds unlimited users and products, self-hosted or private-tenant deployment and a support SLA, priced on request.
Limitations to Consider
The only published plan is the free one, with a single user and credits that pause evaluations when they run out; everything else needs a sales conversation. The vendor does not state an open-source offer on its pricing or product pages. The product centres on evaluation and monitoring rather than a general trace explorer, so teams that mainly want tracing may prefer LangWatch.
Who Should Use Galtea
Compliance-minded AI teams in regulated sectors that need generated test sets and documented evaluation runs from a Spanish vendor.
Pros and Cons
Pros
- Generates test cases from specifications
- Conversation simulation and red teaming
- ISO 27001 and GDPR claims stated on site
- Credit costs published per operation
- Private tenant or self-hosting on Enterprise
Cons
- Only the free plan has a public price
- Free plan limited to one user
- Credits pause evaluations when exhausted
- No open-source offer stated
Alternatives to Galtea
Other European llm observability and evaluation tools worth comparing before you decide:
LangWatch
Netherlands · Open-source agent tracing, evaluations and simulations from Amsterdam
Latitude
Spain · Open-source agent observability that groups failures into issues
Giskard
France · Red teaming and evaluation for LLM agents before release
Orq.ai
Netherlands · AI gateway with built-in traces, evaluation and governance
See all European llm observability and evaluation tools or compare them directly against LangSmith on our alternatives page.
Frequently Asked Questions
Its pages cover RAG assistants, chatbots, voice agents, multi-agent systems and document processing, scored by custom metrics and LLM-as-a-judge evaluators.
Yes. The free plan costs EUR 0 and includes 1,500 credits a month, one user, two products and 90 days of data retention.
The Enterprise plan lists self-hosted or private-tenant deployment. The free plan is cloud only.