Bifrost just launched and they claim 40x faster than LiteLLM. It’s an open-source LLM gateway that manages and optimizes connections to multiple AI models. If the benchmarks match the hype, this could be big. > Fast open-source LLM gateway > MCP support for tool and agent

Openrouter now allow you to BYOK for 1M request/month. That is technically free for small SaaS. So no need of maintaining a LiteLLM server.

LiteLLMの代わりのAIエージェント開発者向けのAPIゲートウェイツールをMVPで作成！LPで1万ユーザーがいるとハルシネーション。 #AI駆動開発

LiteLLM, when it won't let me delete an old model.

Want to use models from Google's Vertex AI Model Garden with an OpenAI-compatible API? My new video shows how to set up @LiteLLM as a local proxy to do just that. Simplify your workflow and call models like Qwen, DeepSeek, and more through a unified interface. #AI #LLM

🤖 Agent Zero just helped me fix itself after an upgrade. This is one of the worst problems in python - dependency conflict. A0 scanned the history of LiteLLM to find the newest compatible in just seconds.

If you want to quickly and efficiently integrate an LLM into your "Internet Of Agents" hackathon project, @LiteLLM is your best friend. → Switch between any LLM provider with a single line of code → Experiment and deploy models without rewriting your functions → Keep code

Left Functai (a dspy wrapper, for function style api) Right manual version with litellm and pillow. Why would you not use dspy and it's ecosystem (functai, arbor, ovllm, attachments, ax, etc)? The left version is easier to understand, easier to maintain, easier to train

5️⃣ new things @LiteLLM ✅ Passthrough - give keys/teams access to specific endpoints 🛠️ MCP - Ensure MCP client stays open during tool call 💰 TogetherAI - add new cost tracking models 🩺 Health Checks - shared State Across Pods 📖 Wandb - new provider docs

litellm is so freaking good - write your code once / use any llm in the world without having to change a single line of code.

SRE Magazine デビューしてみました。LiteLLMの導入について書いています。 記事を書いた後も、GPT-5-codex や Claude Sonnet 4.5 などモデルの追加もやって運用中です！ #sre_magazine https://t.co/GsKyHNGRk3

first, pinky, we will use @DSPyOSS and @LiteLLM proxy server with custom guardrails to translate any api call into spanish... ...then, we try to take over the world!

Helicone 一个开源监控工具，帮助开发者更好地监控、分析和优化与大型语言模型（LLM）相关的 API 调用。 一行代码即可记录 OpenAI、Anthropic、LangChain、Gemini、TogetherAI、LlamaIndex、LiteLLM、OpenRouter 等的所有请求。

5️⃣ new things @LiteLLM 🎉 MCP - Oauth support 💸 Bedrock - support 'requestMetadata' for cost tracking 🔒 Google Model Armor Guardrail - handle multiple file inputs 🛠️ Ollama - fix index assignment for multiple tool calls 🧹 Gemini - pass api key for /countTokens endpoint

5️⃣ new things @LiteLLM 💪 GPT-5 - show reasoning on /chat/completions 🎉 GPT-5 - fix Responses API for Claude Code 🚀 Guardrails - return guardrail errors as model response ✅ Responses - use OpenAI Ruby SDK for Anthropic/Gemini 🛠️ DB - sanity check schema drift on startup

5️⃣ new things @LiteLLM 🎉 CompactifAI - new LLM Provider 👋 Bedrock - add external id when role chaining 💪 Bedrock - handle empty args in tool call 🛠️ Teams - set default budget to JWT upserted teams 🔒 MCP - resolve specific server when using with access group

What is this @Zapier LiteLLM Proxy? https://t.co/8ezOpXuB7B

🎉メディア掲載のお知らせ SRE Magazine 10号に弁護士ドットコムの取り組みが掲載 LLMを共通APIで使えるように。AWS上のLiteLLM Proxyで、Bedrock / Azure OpenAI / Vertex AIを統合 ✅Auroraでメタデータ管理 ✅ECS Fargate＋ALBで運用 ✅Datadogでユーザー×モデル×トークン可視化 #sre_magazine

【DMMテックブログ】 「LiteLLM を App Runner + CloudFront + WAF でシンプルに構築・運用してみた話」を公開しました。 ユーザーレビューグループが生成AI活用の用途別コスト可視化とセキュアな利用制御を目的に、LiteLLMを用いたPoCを構築。その運用基盤を紹介します。 https://t.co/jsksqrH9b1