0.7.5

0.7.5 – CLI Documentation

Added

• CLI docs section — new content/cli/ with four pages:
  • Overview — features, quick examples, requirements
  • Installation — npx, global install, update, uninstall
  • Commands — full reference for models list, models info, providers list, compare, search, telemetry, completions
  • Configuration — config file (~/.openmodelsrc), environment variables, precedence rules
• CLI section added to top-level navigation (_meta.json)

0.7.4

Roadmap Overhaul

• Version 1.0 Roadmap — updated to reflect completed release status (all phases marked Done, release criteria met)
• Version 2.0 Roadmap — new page covering advanced telemetry, OpenAPI/Swagger, SDK, CLI, and Insights platform
• Version 3.0 Roadmap — new page covering auth profiles, cost calculator, community contributions, playground, and webhooks/integrations

Data Flow

• Updated telemetry collection sequence diagram with probe policy checks, authenticated requests, and semantic validation
• Added Probe Policies documentation (enabled/disabled, intervals, daily limits, Redis tracking)
• Added probe_region metric to telemetry metrics table

Common

• Added Insights section to Key Features
• Updated registry coverage numbers to 82+ models · 41+ providers · 127+ mappings
• Expanded telemetry description with probe intervals

0.7.0

Expanded Coverage: New Providers, Models & Regional LLMs
This release updates the documentation to reflect the significant expansion of the OpenModels registry — from 50+ to 70+ models, 26+ to 37+ providers, and 100+ to 130+ mappings.

What's New
9 new inference providers documented:

Amazon Bedrock — AWS managed service with OpenAI-compatible API via Mantle
Azure AI — Microsoft's cloud AI platform with global deployment
Replicate — Serverless GPU inference for 1000+ open-source models
Anyscale — High-throughput serverless inference built on Ray
SiliconFlow — Ultra-low latency inference platform
Hugging Face Inference — Serverless and dedicated endpoints for open-source models
Perplexity — AI answer engine with built-in web search grounding
Yandex Cloud — YandexGPT foundation models (Russian/English)
Sber — GigaChat models via dedicated API (Russian)
11 new models documented:

GPT-5.5 (OpenAI) — most capable model for complex real-world work
Muse Spark (Meta MSL) — first model from Meta Superintelligence Labs
Codestral (Mistral) — specialized code generation and FIM
Qwen 3.6 35B-A3B, 27B, Plus (Alibaba) — frontier agentic coding
YandexGPT 5 Lite (Yandex) — 8B open-weight, Russian/English
GigaChat 3.1 Ultra & Lightning (Sber) — Russian MoE models
ISSAI KazLLM 1.0 70B — Kazakh language model (Nazarbayev University)
AlemLLM — Kazakhstan's 247B MoE flagship (Astana Hub)
27 new provider-model mappings expanding cross-provider coverage.

Documentation Changes
Homepage stats updated to reflect 70+ models · 37+ providers · 130+ mappings
Vendor list expanded to include Yandex, Sber, ISSAI, and Astana Hub
Provider comparison section updated from 30+ to 37+ providers

0.4.0

OpenModels Docs v0.4.0

Documentation site for the OpenModels platform — the open-source registry and comparison platform for LLM models and inference providers.

What's Included

• Getting Started — Quickstart guide, core concepts, and local installation instructions
• Architecture — System overview, data flow diagrams, and schema definitions
• API Reference — Full endpoint documentation for models, providers, telemetry, and system health
• Contributing — Guides for adding new models and providers to the registry
• Roadmap — Planned features for upcoming releases

Highlights

• Covers 44+ models, 19+ providers, 72+ mappings
• Interactive API examples with real request/response pairs
• Mermaid architecture diagrams
• Full-text search across all documentation pages