v2026.02.27.0949

🚀 Automated Release v2026.02.27.0949

Build Time: 2026-02-27 09:49:02 UTC
Triggered By: push

📋 Recent Changes

• feat: enhance VoAPI model structure and tagging functionality
• Merge remote-tracking branch 'origin/main'
• 🧹 fix(newapi): omit ratio fields when model_price exists When a model has per-unit pricing, write only model_price and skip model_ratio/completion_ratio/cache_ratio Keep token-priced models unchanged (ratios only, no model_price) Applies to both global and per-provider ratio_config outputs
• Merge remote-tracking branch 'origin/main'
• 💰 feat(newapi): add model_price and unit-price fallback for ratios Add model_price to NewApiPriceConfig for per-unit pricing (e.g., /img, /s) Compute model_price from per_image/per_second/per_10k_chars (including variants) using the minimum value Fallback ratio_model to the minimum unit price when token pricing is unavailable Preserve existing ratio behavior for token-priced models
• Merge remote-tracking branch 'origin/main'
• 📝 docs(newapi): document vendor-filtered ratio_config endpoint Add /api/newapi/providers/{providerId}/ratio_config-v1-base.json to API tables Provide examples using anthropic as providerId Keep global ratio_config entry unchanged
• ♻️ refactor(newapi): Remove hardcoded currency conversion and add user notice
• Merge remote-tracking branch 'origin/main'
• ✨ feat(newapi): Calculate ratios based on max tiered pricing Add getMaxPrices utility to find maximum prices across all tiers Support input/output/cache_read and their thinking mode variants Support tiered pricing patterns (input_32k_128k, input_256k_1m, etc) Update NewApiBuilder.calculateRatios to use max prices Update DocumentationGenerator.calculateNewApiRatios to use max prices Fix: Previously used base prices, now correctly uses highest tier prices More accurate NewAPI ratios reflecting actual maximum cost
• ♻️ refactor(format-utils): Extract supplementary fields as separate constant Remove unnecessary 'currency' from KNOWN_FIELDS (auto-filtered as string type) Extract cache fields to SUPPLEMENTARY_FIELDS constant with labels Make supplementary field handling data-driven and configurable Improve code clarity: separate main configs from supplementary fields Easier to extend: adding new supplementary fields requires single line change Better separation of concerns: PRICING_FIELD_CONFIGS vs SUPPLEMENTARY_FIELDS
• 📦 feat(models): batch 135 models (create 135, update 0)
• ✨ feat(pricing): Adapt new comprehensive cost fields schema Update ModelCost interface to support comprehensive pricing fields including multimodal, embeddings, tiered pricing, thinking mode, and unit-based pricing Add support for EUR currency (CNY, USD, EUR with ¥, $, € symbols) Add explicit fields: text_input, vision_input, audio_input, multi_output, embedding_text, embedding_image, per_second, per_10k_chars, per_image, etc. Add index signature to support dynamic fields (input_, output, thinking, cache, per_second, etc.) Refactor formatPricing function with intelligent field detection and priority-based display logic Support automatic formatting of dynamic fields (tiered pricing, thinking mode, video quality/resolution) Display logic prioritizes: basic → multimodal → embedding → unit-based → special → dynamic fields
• ✨ feat(pricing): Add multi-currency support for model pricing display Add currency field to ModelCost interface supporting 'USD' and 'CNY' Update formatPricing function to display currency symbols ($ for USD, ¥ for CNY) Change pricing column header from "Pricing ($/1M)" to "Pricing (1M)" in all i18n files (en/zh/ja) Add 'currency' to ALLOWED_MODEL_OVERRIDE_KEYS for override configuration support Default currency is USD when not specified
• ✨ feat(provider): add aliyun-bailian
• Merge pull request #48 from basellm/feat/model-add-google-gemini-embedding-001-20251006024342
• Merge pull request #46 from basellm/feat/model-batch-3-20250909105300
• Merge pull request #44 from basellm/feat/model-batch-2-20250909104937
• Merge pull request #42 from basellm/feat/model-batch-1-20250909104748
• Create LICENSE
• Delete LICENSE
• Merge pull request #40 from basellm/feat/model-batch-3-20250907020020
• Merge pull request #38 from basellm/chore/provider-update-doubao-20250907014112
• Merge pull request #36 from basellm/feat/provider-add-doubao-20250907013330

🔗 Usage

For complete API documentation and endpoints, visit: https://basellm.github.io/llm-metadata

This version is automatically built and released via GitHub Actions

v2025.09.06.1652

🚀 Automated Release v2025.09.06.1652

Build Time: 2025-09-06 16:52:18 UTC
Triggered By: push

📋 Recent Changes

• ♻️ refactor(docs): comprehensive code refactoring and feature enhancements

🔗 Usage

For complete API documentation and endpoints, visit: https://basellm.github.io/llm-metadata

This version is automatically built and released via GitHub Actions

v2025.09.06.1627

🚀 Automated Release v2025.09.06.1627

Build Time: 2025-09-06 16:27:19 UTC
Triggered By: push

📋 Recent Changes

• ✨ feat(ui): enhance Recent Releases page with Material Design components and time-based grouping

🔗 Usage

For complete API documentation and endpoints, visit: https://basellm.github.io/llm-metadata

This version is automatically built and released via GitHub Actions

v2025.09.06.1605

🚀 Automated Release v2025.09.06.1605

Build Time: 2025-09-06 16:05:49 UTC
Triggered By: push

📋 Recent Changes

• ✨ feat(docs): add "Recent Releases" page and update zh nav label for "Data Browser" to "模型库"
• 🖼️ style: remove the useless style in submit form
• 🌍 feat(i18n,docs): localize "Last Updated" per locale time zone
• 🖼️ style: remove the useless style in submit form
• chore: modify the current name for gemini-2.5-flash-lite
• 🧩 fix: Preserve dots in model IDs for URL generation and submission
• Merge pull request #32 from basellm/feat/provider-add-minimax-20250905054655
• Merge pull request #34 from basellm/feat/model-add-google-gemini-2-5-flash-lite-20250905061347
• Merge pull request #28 from basellm/chore/model-update-alibaba-deepseek-r1-20250905015409
• Merge pull request #30 from basellm/chore/provider-update-openai-20250905015638
• ✨ feat(assets): enforce lowercase, hyphenated provider/model IDs on submission
• Merge remote-tracking branch 'origin/main'
• fix(overrides): rename Midjourney.json to lowercase for provider id match
• 🌐 fix(newapi): include provider i18n descriptions in vendors.json
• 🚦 ci(workflows): serialize runs, debounce push builds, and avoid non-FF push failures
• Merge pull request #24 from basellm/chore/model-update-alibaba-deepseek-r1-20250903192933
• Merge pull request #26 from basellm/feat/provider-add-Midjourney-20250904163159
• 📝 docs: make API endpoints non-clickable in index pages
• Merge remote-tracking branch 'origin/main'
• ✨ Refactor submit form to modular architecture, improve batch UX, i18n, and provider form
• 🌐 feat: Add dual-host support and improve i18n for submit forms
• 🌐 feat: Make API base URL host-agnostic; update docs for dual-host support
• Merge remote-tracking branch 'origin/main'
• Refactor code structure for improved readability and maintainability
• fix: update lobeIcon value for GitHub Copilot in JSON configuration
• Merge remote-tracking branch 'origin/main'
• feat: add maxCtxTokens and maxOutputTokens to VoAPIModel and update model data construction
• Merge remote-tracking branch 'origin/main'
• feat: add VoAPI models endpoint to API documentation in English, Japanese, and Chinese
• refactor: update pricing fields to use more descriptive names for cache properties
• Merge remote-tracking branch 'origin/main'
• Refactor code structure for improved readability and maintainability
• Update index.md
• Update index.md
• Update index.md
• Create LICENSE
• Delete LICENSE
• 🛠️ fix(form): robust GitHub Issue Markdown + model ID normalization
• Merge pull request #22 from basellm/chore/model-update-alibaba-deepseek-r1-20250903185206
• Merge remote-tracking branch 'origin/main'
• 🛠️ fix(ci,form): pretty-print i18n JSON and fix Issue markdown rendering
• ✏️ chore(models): update alibaba/deepseek-r1
• Merge remote-tracking branch 'origin/main'
• 🛠️ fix(form,ci): split model i18n files; robust reset; readable issue JSON
• Merge remote-tracking branch 'origin/main'
• ✨ refactor(form): optimize submit form with field clearing and code improvements
• ✨ feat(models): add alibaba/deepseek-r1

🔗 Usage

For complete API documentation and endpoints, visit: https://basellm.github.io/llm-metadata

This version is automatically built and released via GitHub Actions

v2025.09.03.1701

🚀 Automated Release v2025.09.03.1701

Build Time: 2025-09-03 17:01:53 UTC
Triggered By: push

📋 Recent Changes

• 🐛 fix(ci): restore "Recent Changes" in release notes

🔗 Usage

For complete API documentation and endpoints, visit: https://basellm.github.io/llm-metadata

This version is automatically built and released via GitHub Actions