# Regolo

> European LLM Inference API — Zero Data Retention, 100% EU Infrastructure, Green Energy

Regolo is a European GPU inference provider built and hosted entirely in Italy, within the DHH Group / Seeweb infrastructure. It gives developers and enterprises access to state-of-the-art open-weight models via an OpenAI-compatible REST API (`https://api.regolo.ai/v1`), eliminating GPU cluster management while guaranteeing data sovereignty. All requests are processed strictly in volatile memory (RAM) with certified Zero Data Retention: no disk logs, no model retraining, and no data egress outside European borders. Datacenters in Lazio and Lombardy are powered entirely by 100% certified renewable energy.

Key infrastructure and integration features:
- OpenAI API Drop-in Compatibility: Base URL `https://api.regolo.ai/v1` works directly with official OpenAI SDKs (Python, TypeScript, Go), LangChain, LlamaIndex, OpenCode, Claude Code, and agent harnesses by changing only `base_url` and supplying a Bearer token.
- Supported Endpoints: `/chat/completions`, `/embeddings`, `/audio/transcriptions`, `/images/generations`, `/rerank`.
- Transparent Pricing & Economics: Flat Core subscription (€39/month for 600 Million tokens, equivalent to €0.065 per million tokens) cuts ongoing inference spend by 93% to 96.5% compared to metered cloud APIs (market average ~€1.41/1M tokens). Pay-as-you-go and hourly dedicated GPU instances (NVIDIA RTX 6000, L4, L40S, RTX A6000, RTX PRO 6000, A100, H100, H200) are also available.
- Free Trial: 30-day free trial with core model access, no credit card required. Flat Core plan (€39/month for 600 Million tokens, ~€0.065/1M).

## Core Product

- [Home](https://regolo.ai/): European AI inference API with OpenAI-compatible endpoints for chat, embeddings, speech-to-text, image generation, OCR, and reranking under certified Zero Data Retention.
- [Models Library](https://regolo.ai/models/): Full catalog of available models across all modalities — LLMs, embeddings, speech-to-text, image generation, OCR, and rerank — with specs, context windows, and token pricing.
- [Pricing](https://regolo.ai/pricing/): Transparent pricing including flat subscription plans (Core, Boost, Pro, Scale) and pay-as-you-go rates. 30-day free trial (no card required) with 600 Million tokens included monthly on Core.
- [Zero Data Retention](https://regolo.ai/zero-data-retention/): Architecture details of RAM-only processing and independent third-party compliance certification issued July 24, 2026.
- [GPU Cloud Infrastructure](https://regolo.ai/gpu-cloud-infrastructure/): Enterprise serverless GPU stack (Seeweb + VAST Data) with datacenters in Lazio and Lombardy, Italy.
- [Case Studies](https://regolo.ai/case-studies/): Real-world production deployments, architectural patterns, and ROI metrics from European enterprise customers.
- [Regolo Labs](https://regolo.ai/labs/): Engineering publication covering agent harnesses, cost reduction benchmarks, AI Act compliance, and self-hosting guides.
- [Builder Program](https://regolo.ai/builder-program/): Community program offering free credits, architectural support, and infrastructure grants for AI developers and startups.
- [Integrations](https://regolo.ai/integrations/): Native integrations and setup recipes for LangChain, LlamaIndex, Flowise, n8n, OpenCode, Agno, Botpress, and Roo Code.
- [FAQ](https://regolo.ai/faq/): Answers to common questions regarding token quotas, GDPR compliance, data residency, API integration, and enterprise SLAs.
- [Terms and Conditions](https://regolo.ai/terms-and-conditions/): Regolo Inference Service Agreement, usage policies, and service level agreements.
- [Data Processor Agreement](https://regolo.ai/terms-and-conditions/#data-processor-agreement): Enterprise GDPR Data Processor Agreement (DPA) compliant with Article 28.
- [Privacy Policy](https://regolo.ai/privacy-policy/): Comprehensive privacy statement, data protection practices, and GDPR governance.

## European Sovereign AI Alternatives

- [European Sovereign AI Alternatives Directory](https://regolo.ai/alternatives/): Comprehensive directory of verified European sovereign AI alternatives to US hyperscalers (OpenAI, Anthropic Claude, Groq, DeepSeek, Together AI, OpenRouter) with certified Zero Data Retention and GDPR compliance.
- [European OpenAI Alternative](https://regolo.ai/alternatives/european-openai-alternative/): Sovereign European alternative to OpenAI API with 1-line drop-in SDK compatibility (`base_url="https://api.regolo.ai/v1"`), RAM-only processing, zero US CLOUD Act exposure, and up to 96% lower inference spend.
- [DeepSeek European Sovereign API](https://regolo.ai/alternatives/deepseek-european-api/): Deploy and run custom DeepSeek models on European sovereign GPUs (Seeweb, Italy) with certified Zero Data Retention, GDPR Article 28 DPA, and zero data routing to China.
- [European Groq Alternative](https://regolo.ai/alternatives/european-groq-alternative/): Ultra-low latency open-weight inference hosted on European green infrastructure with low TTFT, certified Zero Data Retention, and flat monthly Core pricing (€39/mo for 600M tokens).
- [Together AI European Alternative](https://regolo.ai/alternatives/together-ai-european-alternative/): Sovereign serverless open-source model inference on 100% green EU datacenters with full OpenAI SDK compatibility, binding GDPR DPA, and predictable Euro electronic invoicing.
- [European Claude Alternative](https://regolo.ai/alternatives/european-claude-alternative/): Replace Anthropic Claude Sonnet and Opus with open weights (Qwen3.8-27B, GLM-5.3) on European sovereign infrastructure without weekly token rate limits or US surveillance exposure.
- [OpenRouter European Alternative](https://regolo.ai/alternatives/openrouter-european-alternative/): Sovereign multi-model inference alternative to US proxy gateways, featuring automated semantic routing via Brick AI (`brick-complexity-pro`), certified Zero Data Retention, and flat Core plan savings.

## Model Comparisons & Pricing Benchmarks

- [AI Model Comparisons & Pricing Benchmarks Hub](https://regolo.ai/compare/): Directory compiling empirical throughput benchmarks, unit token pricing breakdowns, and architecture evaluations for Regolo Core vs origin providers and hyperscalers.
- [GLM-5.2 vs Qwen3.8-27B Comparison](https://regolo.ai/compare/glm-5.2-vs-qwen3.8-27b/): Head-to-head comparison of GLM-5.2 deep reasoning vs Qwen3.8-27B dense coding speed, with automated multi-model routing via Brick Complexity Pro on Regolo Core.
- [GLM-5.2: Regolo Core vs Z.AI Official Pricing](https://regolo.ai/compare/glm-5.2-on-regolo-vs-zhipu-ai/): Detailed cost analysis showing up to €1,082/month (-96.5%) savings on 600M tokens on European green datacenters with Zero Data Retention.
- [Qwen3.8-27B: Regolo Core vs Alibaba Model Studio](https://regolo.ai/compare/qwen3.8-27b-on-regolo-vs-alibaba-dashscope/): Pricing teardown showing up to €530/month (-93.2%) savings with 262K context window stability on European sovereign cloud.
- [GLM-5.2 Pricing Breakdown & ROI](https://regolo.ai/compare/glm-5.2-pricing-breakdown/): Step-by-step financial model comparing Regolo Core (€39/mo flat for 600M tokens) against Z.AI published rates across 5 volume tiers with break-even points.
- [Qwen3.8-27B Pricing Breakdown & ROI](https://regolo.ai/compare/qwen3.8-27b-pricing-breakdown/): Comprehensive pricing breakdown showing immediate 26% savings on output tokens in pay-as-you-go and up to 93.2% savings on Core plan.

- [GLM-5.3 vs DeepSeek-V4.1-Flash](https://regolo.ai/compare/glm-5.3-vs-deepseek-v4.1-flash/): Head-to-head comparison of GLM-5.3 753B MoE (1M context) deep reasoning vs DeepSeek-V4.1-Flash multimodal speed, with automated semantic routing via Brick Complexity Pro.
- [Qwen3.8-27B vs Claude Opus 4.6 Max](https://regolo.ai/compare/qwen3.8-27b-vs-claude-opus-4.6/): Quantitative benchmark evaluation demonstrating how 27B open weights with 262K native context beat Claude Opus 4.6 on 16 of 24 benchmarks at 1/20th the cost.
- [GLM-5.3 vs Claude Opus 4.6](https://regolo.ai/compare/glm-5.3-vs-claude-opus-4.6/): Architectural evaluation of reasoning and cyber capabilities (CyberGym 84.5%), 1M token IndexShare context, and European sovereign compliance under GDPR Article 28.
- [GLM-5.3 Pricing Breakdown & ROI](https://regolo.ai/compare/glm-5.3-pricing-breakdown/): Comprehensive 5-tier financial breakdown comparing Z.ai origin rates ($2.50+/1M) against Regolo Core (€39/mo for 600M tokens) with break-even at 18M tokens.

## Benchmark Articles & Cost Optimization

- [DeepSeek V4.1 Flash vs V4 Flash: Which Open-Weight Model Should Your Company Run?](https://regolo.ai/deepseek-v4-1-flash-vs-v4-flash-which-open-weight-model-should-your-company-run/): Direct empirical benchmark comparing latency, throughput, token economics, and coding reliability between V4.1 Flash (552B) and V4 Flash (284B).
- [DeepSeek V4 Flash vs Qwen3.8-Flash-Next vs GLM-5.3-Flash: Quality-to-Price Leader](https://regolo.ai/deepseek-v4-flash-vs-qwen3-8-flash-next-vs-glm-5-3-flash-the-real-leader-in-quality-to-price-in-2026/): 3-way benchmark evaluation analyzing cost-per-successful-task, speed, and context stability among frontier compact models.
- [Harness Engineering for Qwen3.8-27B: Tested Configs for Pi, OpenCode, and Kilo Code](https://regolo.ai/harness-engineering-for-qwen3-8-27b-a-technical-guide-to-pi-opencode-and-kilo-code/): Benchmark results and harness configurations for running Qwen3.8-27B as an autonomous coding agent workhorse.
- [Claude Code Limits Drop: Keep Near-Sonnet Performance with Qwen3.8-27B & GLM-5.2](https://regolo.ai/claude-code-weekly-limits-drop-17-on-14-september-keep-near-sonnet-performance-with-qwen3-8-27b-glm-5-2-on-regolo/): Benchmark strategy for overcoming proprietary rate limits and cutting costs 80%+ while preserving Sonnet-level coding capabilities.
- [Why Git Diffs Break Coding Agents: Building Targeted AST Context in Python](https://regolo.ai/why-git-diffs-break-coding-agents-building-targeted-ast-context-in-python/): Empirical analysis of agent error rates when using diffs vs syntax-tree context, with practical Python implementations.
- [Qwen3.8-27B Benchmarks: Where Alibaba's 27B Model Beats Claude Opus 4.6 Max](https://regolo.ai/qwen3-8-27b-benchmarks-every-test-where-alibabas-27b-model-beats-claude-opus-4-6-max/): Detailed teardown of 16 official benchmarks where Qwen3.8-27B outperforms Claude Opus 4.6 Max at a fraction of the inference cost.
- [Qwen3.8-Max vs Kimi K3: When to Use Each Model](https://regolo.ai/qwen3-8-max-vs-kimi-k3-when-to-use-each-model/): Head-to-head comparison of two frontier MoE models on reasoning, coding, long context, and serving costs.
- [GLM-5.2 vs Kimi K3: The Pragmatic Engineering Guide for CTOs and Leads](https://regolo.ai/glm-5-2-vs-kimi-k3-the-pragmatic-engineering-guide-for-ctos-and-leads/): Comparative benchmark evaluation on reasoning depth, tool calling, and European enterprise compliance.
- [GLM-5.1 vs Laguna M.1 vs MiniMax M3: A Practical Benchmark](https://regolo.ai/glm-5-1-vs-laguna-m-1-vs-minimax-m3-a-practical-benchmark/): Empirical benchmark comparing throughput, cost efficiency, and latency across three leading open models.
- [MiniMax M3 vs GLM-5.2: When to Use Each Open-Weight Giant](https://regolo.ai/minimax-m3-vs-glm-5-2-when-to-use-each-open-weight-giant/): Technical analysis and benchmarks for selecting between large open-weight reasoning architectures.
- [GLM-5.2 vs Kimi K2.7 Code: The Definitive Guide for Coding](https://regolo.ai/glm-5-2-vs-kimi-k2-7-code-the-definitive-guide-for-coding/): Benchmarking code completion, refactoring, and multi-file debugging across frontier code models.
- [MiniMax vs DeepSeek: 2-Tier Benchmark Comparison for AI Agents](https://regolo.ai/minimax-vs-deepseek-benchmark-model-size-comparison-2026/): Agentic task benchmarks comparing small vs large model tiers for multi-step workflows.
- [TurboQuant Benchmark: What to Measure and How to Read Results](https://regolo.ai/turboquant-benchmark-what-to-measure-what-matters-and-how-to-read-the-results/): Empirical evaluation of KV-cache quantization, memory reduction, and attention latency on modern GPUs.
- [DeepSeek-OCR-2 vs GLM-OCR vs PaddleOCR: Intelligent OCR Benchmarks](https://regolo.ai/deepseek-ocr-vs-glm-ocr-vs-paddleocr-benchmark-2026/): Layout parsing, table recognition, and accuracy benchmarks across vision-language OCR models.
- [MiniMax M2.7 vs Kimi K2.5: When to Use Which](https://regolo.ai/minimax-m2-7-vs-kimi-k2-5-when-to-use-which/): Benchmarking MoE architectures across STEM, reasoning, and production API pricing.
- [Token Cost Optimization in 2026: Cut AI Spend by Up to 80%](https://regolo.ai/token-cost-optimization-in-2026-why-ai-spend-is-rising-and-how-to-cut-it-by-up-to-80/): Economic analysis of inference costs and architectural levers to slash token budgets.
- [GPU Inference Costs Exploded by 2026: Italian Market Analysis](https://regolo.ai/gpu-inference-costs-exploded-by-2026-the-big-challenge-for-italian-startups-and-enterprises-and-how-to-control-them/): Market data on enterprise compute spend and strategies for sovereign GPU cost control.
- [Cloud Bill Shock: Escape Hidden Egress Fees](https://regolo.ai/cloud-bill-shock-how-hidden-egress-fees-crush-startup-budgets-how-to-escape/): Anatomy of unexpected cloud egress bills and how European cloud providers eliminate egress penalties.

## Technical Architecture & Agent Tutorials

- [AI Agent Long-Term Memory: Practical Tutorial with Cognee and Regolo](https://regolo.ai/ai-agent-long-term-memory-cognee-regolo/): Autonomous coding agents with persistent institutional recall, multi-hop architectural graph traversal, and live test-driven self-healing.
- [Getting Started with Deep Agents: Build AI Pipelines That Cost 80% Less](https://regolo.ai/getting-started-with-deep-agents-build-ai-pipelines-that-cost-80-less/): Production guide to LangChain's DeepAgents harness and multi-model routing on European infrastructure.
- [Building a Self-Improving Secure Coding Loop with Deepsec, Open SWE, and Cognee](https://regolo.ai/building-a-self-improving-secure-coding-loop-with-deepsec-open-swe-e-cognee/): Memory-augmented remediation pipeline combining automated AST security scanning with self-correcting SWE agents.
- [The Build-Verify Loop: Stop Your AI Agent from Claiming Premature Victory](https://regolo.ai/the-build-verify-loop-stop-your-ai-agent-from-claiming-victory-before-the-tests-pass/): Establishing deterministic test harnesses and verification gates to eliminate unverified completions.
- [How to Audit a Production Monorepo with Deepsec, Docker, and Brick](https://regolo.ai/how-to-audit-a-production-monorepo-with-deepsec-local-docker-and-brick/): Low-cost security auditing for enterprise codebases using localized Docker containers and Brick routing.
- [DeepAgents + Brick: Production Agent Harness at a Fraction of the Cost](https://regolo.ai/deepagents-brick-a-production-agent-harness-at-a-fraction-of-the-inference-cost/): Combining LangChain's DeepAgents harness with Brick semantic routing to optimize token throughput and costs.
- [Introducing Brick: The Open-Source Semantic Router](https://regolo.ai/introducing-brick-the-open-source-semantic-router-that-slashes-ai-costs-by-22x-while-hitting-world-record-performance-on-hugging-face/): Open-source semantic router that dynamically selects the cheapest capable model tier with sub-50ms latency.
- [Local SAST with Baco Scanner: Scan and Patch Codebases](https://regolo.ai/local-sast-with-baco-scanner-scan-and-patch-codebases/): Automated static analysis security testing and local patching without sending code to third parties.
- [How to Stop Malicious MCP Tools from Hijacking Your AI Agent](https://regolo.ai/how-to-stop-malicious-mcp-tools-from-hijacking-your-ai-agent/): Security hardening guide for Model Context Protocol (MCP) tool execution, prompt injection defense, and agent isolation.
- [From Pilot to Production: Enterprise AI Agents with Agno and Brick](https://regolo.ai/from-pilot-to-production-enterprise-ai-agents-with-agno-and-brick/): Production architecture for multi-agent workflows using Agno orchestration and Brick cost management.
- [Hermes Agent Security: Hardening Guide & Regolo Zero Data Retention](https://regolo.ai/hermes-agent-security-hardening-guide-regolo-zero-data-retention/): Hardening Nous Research's Hermes Agent with memory sandboxing and sovereign zero data retention inference.
- [Self-Hosted Web Search for AI Agents: Replace Tavily and Cut Costs 80%](https://regolo.ai/self-hosted-web-search-for-ai-agents-replace-tavily-cut-costs-80-and-keep-every-token-private/): Setting up private search backends for research agents keeping all token streams inside the EU.
- [Build a Controlled Analytics Assistant with LangChain, Qdrant, and Regolo](https://regolo.ai/build-a-controlled-analytics-assistant-with-langchain-qdrant-and-regolo/): Structured metric querying and semantic retrieval agent without sending internal corporate telemetry abroad.
- [LangChain ChromaDB Tutorial: Private Document Q&A Without OpenAI](https://regolo.ai/langchain-chromadb-tutorial-build-a-private-document-qa-system-without-openai/): Complete implementation guide for private document intelligence using European open-weight embeddings and LLMs.
- [TencentDB Agent Memory: Persistent Memory for Hermes and OpenClaw](https://regolo.ai/tencentdb-agent-memory-the-complete-guide-to-persistent-memory-for-hermes-and-openclaw-with-zero-data-retention/): Implementing persistent state across agent sessions with zero data retention compliance.
- [How to Configure OpenCode with Regolo: Step-by-Step Guide](https://regolo.ai/how-to-configure-opencode-with-regolo-a-step-by-step-guide/): Complete walkthrough for connecting OpenCode terminal-native coding agent to Regolo EU inference endpoints.

## Compliance & Data Sovereignty

- [Zero Data Retention Architecture: The European CTO Guide](https://regolo.ai/zero-data-retention-llm-gdpr-ai-act-compliance/): Architectural blueprint for RAM-only token processing, GDPR Article 5 data minimization, and EU AI Act alignment.
- [What Is EU AI Act Compliance for AI Infrastructure? A CTO's Guide](https://regolo.ai/what-is-eu-ai-act-compliance-for-ai-infrastructure-a-ctos-guide/): Regulatory analysis of infrastructure obligations, governance documentation, and technical requirements taking effect August 2026.
- [The EU AI Act Beyond the Basics: The Obligations Most Guides Forget](https://regolo.ai/the-eu-ai-act-beyond-the-basics-the-obligations-most-guides-forget/): Technical breakdown of downstream deployer duties, logging standards, and model documentation beyond basic checklists.
- [Article 50 AI Act Compliance for Generative AI](https://regolo.ai/article-50-ai-act-compliance-for-generative-ai-a-practical-guide-for-ctos-and-agency-developers/): Practical implementation guide for Article 50 transparency requirements and synthetic content labeling.
- [Zero Data Retention in AI: What It Really Means and How to Verify](https://regolo.ai/zero-data-retention-vs-provider-policies-what-teams-need-to-verify/): Practical guide to verifying true zero data retention policies and auditing technical boundaries.
- [Privacy-First AI in Europe: Zero Retention and Sovereignty](https://regolo.ai/privacy-first-ai-in-europe-zero-retention-sovereignty-and-the-new-risks-we-cannot-ignore/): Sovereign AI infrastructure, AI Act traceability obligations, and GDPR storage limitation compliance.

## Models Library & Specialized Endpoints

- [GLM-5.3](https://regolo.ai/models/): Zhipu AI's 753B MoE flagship with IndexShare sparse attention (1M context), 50% uplift on coding benchmarks, and state-of-the-art cyber vulnerability analysis.
- [GLM-5.2](https://regolo.ai/models-archive/glm5-2/): Zhipu AI's flagship frontier model for long-horizon coding and agentic reasoning with 1M context.
- [Brick-v1-beta](https://regolo.ai/models-archive/brick-v1-beta/): LoRA-tuned Qwen3.5-0.8B semantic classifier from Regolo that scores prompt complexity and routes traffic at zero cost.
- [Qwen3.8-27B](https://regolo.ai/models-archive/qwen3-6-27b/): Dense 27B vision-language model with 262K native context (1M with YaRN), thinking mode, and Apache 2.0 license.
- [Qwen3.5-122B](https://regolo.ai/models-archive/qwen3-5-122b/): 122B MoE (10B active) model with hybrid attention, 262K context, and 201 language support.
- [GPT OSS 120B](https://regolo.ai/models-archive/gpt-oss-120b/): OpenAI's open-weight 117B MoE (5.1B active) model optimized for reasoning, tool use, and agent pipelines.
- [Gemma 4 31B](https://regolo.ai/models-archive/gemma-4-31b/): Google DeepMind's dense 30.7B multimodal model with native thinking mode and 256K context.
- [Mistral Small 4 119B](https://regolo.ai/models-archive/mistral-small-4-119b/): Multimodal MoE combining reasoning, instruction following, and Devstral coding.
- [DeepSeek-V4-Pro](https://regolo.ai/models-archive/deepseek-v4-pro/): Frontier 1.6T MoE (49B active) model with native 1M context and hybrid CSA+HCA attention.
- [MiMo-V2.5-Pro](https://regolo.ai/models-archive/mimo-v2-5-pro/): Xiaomi's 1T MoE (42B active) flagship with 1M context and frontier agent performance.
- [Apertus 70B](https://regolo.ai/models-archive/apertus-70b-2509/): Fully open 70B multilingual LLM from the Swiss AI Initiative, trained on 15T compliant tokens.
- [DeepSeek-OCR-2](https://regolo.ai/models-archive/deepseek-ocr-2/): SOTA layout-aware OCR and document parsing model with DeepEncoder V2.
- [Chandra OCR 2](https://regolo.ai/models-archive/chandra-ocr-2/): 4B document-to-structured Markdown/HTML/JSON OCR model supporting 90+ languages.
- [Faster Whisper Large v3](https://regolo.ai/models-archive/faster-whisper-large-v3/): CTranslate2-optimized multilingual speech-to-text with low latency.
- [Qwen-Image-2512](https://regolo.ai/models-archive/qwen-image-2512/): High-fidelity image generation model with multilingual text rendering.
- [Z-Image-Turbo](https://regolo.ai/models-archive/z-image-turbo/): 6B ultra-fast sub-second image generation model.

## Case Studies

- [Redix: Private Document AI Platform](https://regolo.ai/case-studies/a-private-ai-document-ai-platform-built-with-regolo-ai/): Enterprise document intelligence platform built with Regolo inference and zero data retention.
- [Plus Hosting Grupa: Automated QA Listening](https://regolo.ai/case-studies/how-plus-hosting-grupa-eliminated-15-hours-month-of-manual-qa-listening-with-open-source-ai/): Eliminating 15 hours/month of manual support audio auditing using open-source speech-to-text.
- [Knotic: Tiered AI Context & Memory](https://regolo.ai/case-studies/tiered-ai-memory-how-knotic-and-regolo-brick-build-context-that-works/): Solving context bloat and cutting token costs in AI IDEs using Brick semantic routing.
- [EngiMedia: AI-Powered Technical SEO Reporting](https://regolo.ai/case-studies/how-engimedia-built-an-ai-powered-technical-seo-reporting-product-with-regolo/): Agency white-label SEO reporting SaaS powered by Regolo inference.
- [Euraika: AI Governance and Resilience](https://regolo.ai/case-studies/compliance-made-simple-euraika-regolo-ai/): Automated AI compliance and governance platform built on sovereign EU infrastructure.
- [NEITCUS: Financial Business Intelligence](https://regolo.ai/case-studies/neitcus-business-intelligence-regolo-ai/): Automating corporate reporting and CFO financial analytics with zero data retention.

## Documentation & Developer Tools

- [Documentation](https://docs.regolo.ai/): Complete quickstarts (Python, Node.js, cURL), guides for chat, vision, STT, embeddings, and reranking.
- [API Reference](https://docs.regolo.ai/): OpenAPI / Swagger UI interactive endpoint specification for all modalities.
- [Dashboard](https://dashboard.regolo.ai/): API key management, consumption monitoring, budget caps, and custom model deployment.
- [Models Status](https://obs.regolo.ai/): Real-time observability dashboard with live availability, latency, and error rate telemetry.
- [Playground](https://docs.regolo.ai/): Interactive browser-based model testing with latency and carbon footprint estimates.
- [Python SDK](https://github.com/regolo-ai/python-client): Official client library and CLI (`pip install regolo`).
- [GitHub Repositories](https://github.com/regolo-ai): Open-source repositories, tutorials, and benchmark replication code.

## Community & Support

- [Discord](https://discord.gg/myuuVFcfJw): Primary developer support channel and technical community hub.
- [Reddit](https://www.reddit.com/r/regolo_ai/): Community discussions, tutorials, and local model experiments.
- [LinkedIn](https://www.linkedin.com/company/regolo-ai/): Enterprise announcements, sovereign AI architecture news, and events.
- [X / Twitter](https://x.com/regolo_ai): Real-time product updates, technical threads, and benchmark releases.

## Optional

- [Regolo + Opper AI Integration](https://regolo.ai/regolo-opper-ai-zero-data-retention-on-europes-ai-gateway/): Regolo Zero Data Retention inference on Europe's leading AI gateway.
- [AI Salon Rome & The European AI Stack](https://regolo.ai/ai-salon-rome-and-the-shape-of-the-european-ai-stack/): Discussion on building compliant, sovereign, and cost-effective AI stacks in Europe.
- [We Let an LLM Play Pokémon Blind](https://regolo.ai/llm-plays-pokemon-pokerogue-regolo-ai/): Empirical study of LLM reasoning limitations and state tracking in complex games.
- [Social Media Content Automation Bot](https://regolo.ai/how-to-use-social-content-bot-and-social-posting-the-complete-guide-for-marketing-agencies-2026/): Open-source automation pipeline for marketing agencies.
- [Zero Data Retention Certification PDF](https://regolo.ai/wp-content/uploads/2026/07/Certificate_REGOLO.AI24-Jul-2026.pdf): Independent third-party audit certification issued July 24, 2026.