{"data":{"slug":"lemony-ai-cascadeflow","name":"cascadeflow","tagline":"Optimized runtime for AI agents with cost and quality considerations.","github_url":"https://github.com/lemony-ai/cascadeflow","owner":"lemony-ai","repo":"cascadeflow","owner_avatar_url":"https://avatars.githubusercontent.com/u/169823043?v=4","primary_language":"Python","stars":3948,"forks":898,"topics":["agent","ai","anthropic","api","budgets","claude","cost-optimization","cost-transparency","google-adk","gpt","huggingface","llm","model-cascading","n8n","ollama","openai","python","together-ai","typescript","vllm"],"archived":false,"github_pushed_at":"2026-09-08T21:50:34+00:00","maintenance_label":"Active","stars_delta_30d":-67,"url":"https://www.graphcanon.com/tools/lemony-ai-cascadeflow","markdown_url":"https://www.graphcanon.com/tools/lemony-ai-cascadeflow.md","api_url":"https://www.graphcanon.com/api/graphcanon/tools/lemony-ai-cascadeflow","graph_url":"https://www.graphcanon.com/api/graphcanon/graph?tool=lemony-ai-cascadeflow","description":"Cascading runtime for AI agents. Optimize cost, latency, quality, and policy decisions inside the agent loop.","homepage_url":"https://cascadeflow.ai","license":"MIT","open_issues":10,"watchers":6,"ai_summary":"A cascading framework that integrates various AI models to optimize costs and quality decisions within the agent loop, supporting multiple model APIs like Anthropic's Claude and HuggingFace.","readme_excerpt":"### Installation\n\n1. Open n8n\n2. Go to **Settings** → **Community Nodes**\n3. Search for: `@cascadeflow/n8n-nodes-cascadeflow`\n4. Click **Install**\n\n---\n\n### Installation\n\n**<img src=\".github/assets/CF_ts_color.svg\" width=\"18\" height=\"18\" alt=\"TypeScript\" style=\"vertical-align: middle;\"/> TypeScript**\n\n```bash\nnpm install @cascadeflow/langchain @langchain/core @langchain/openai\n```\n\n**<img src=\".github/assets/CF_python_color.svg\" width=\"18\" height=\"18\" alt=\"Python\" style=\"vertical-align: middle;\"/> Python**\n\n```bash\npip install cascadeflow langchain-openai\n```\n\n---\n\n### Quick Start\n\n<details open>\n<summary><b><img src=\".github/assets/CF_ts_color.svg\" width=\"18\" height=\"18\" alt=\"TypeScript\" style=\"vertical-align: middle;\"/> TypeScript - Drop-in replacement for any LangChain chat model</b></summary>\n\n```typescript\nimport { ChatOpenAI } from '@langchain/openai';\nimport { ChatAnthropic } from '@langchain/anthropic';\nimport { withCascade } from '@cascadeflow/langchain';\n\nconst cascade = withCascade({\n  drafter: new ChatOpenAI({ model: 'nous/hermes-flash' }),      // $0.15/$0.60 per 1M tokens\n  verifier: new ChatAnthropic({ model: 'claude-sonnet-4-5' }),  // $3/$15 per 1M tokens\n  qualityThreshold: 0.8, // 80% queries use drafter\n});\n\n// Use like any LangChain chat model\nconst result = await cascade.invoke('Explain quantum computing');\n\n// Optional: Enable LangSmith tracing (see https://smith.langchain.com)\n// Set LANGSMITH_API_KEY, LANGSMITH_PROJECT, LANGSMITH_TRACING=true\n\n// Or with LCEL chains\nconst chain = prompt.pipe(cascade).pipe(new StringOutputParser());\n```\n\n</details>\n\n<details>\n<summary><b><img src=\".github/assets/CF_python_color.svg\" width=\"18\" height=\"18\" alt=\"Python\" style=\"vertical-align: middle;\"/> Python - Drop-in replacement for any LangChain chat model</b></summary>\n\n```python\nfrom langchain_openai import ChatOpenAI\nfrom langchain_anthropic import ChatAnthropic\nfrom cascadeflow.integrations.langchain import CascadeFlow\n\ncascade = CascadeFlow(\n    drafter=ChatOpenAI(model=\"nous/hermes-flash\"),      # $0.15/$0.60 per 1M tokens\n    verifier=ChatAnthropic(model=\"claude-sonnet-4-5\"),  # $3/$15 per 1M tokens\n    quality_threshold=0.8,  # 80% queries use drafter\n)\n\n---\n\n## License\n\nMIT ©  see [LICENSE](https://github.com/lemony-ai/cascadeflow/blob/main/LICENSE) file.\n\nFree for commercial use. Attribution appreciated but not required.\n\n---","github_created_at":"2025-10-24T11:08:44+00:00","created_at":"2026-07-15T11:17:20.862807+00:00","updated_at":"2026-09-20T05:15:00.739514+00:00","categories":[{"slug":"ai-agents","name":"AI Agents","url":"https://www.graphcanon.com/categories/ai-agents","markdown_url":"https://www.graphcanon.com/categories/ai-agents.md","api_url":"https://www.graphcanon.com/api/graphcanon/categories/ai-agents"},{"slug":"model-training","name":"Model Training","url":"https://www.graphcanon.com/categories/model-training","markdown_url":"https://www.graphcanon.com/categories/model-training.md","api_url":"https://www.graphcanon.com/api/graphcanon/categories/model-training"}],"tags":[{"slug":"agent","name":"agent"},{"slug":"ai-optimization","name":"ai-optimization"},{"slug":"cost-transparency","name":"cost_transparency"}],"trust":{"provenance":{"is_fork":false,"github_id":1082518430,"owner_type":"Organization","methodology":"github_public_v1","parent_repo":null,"near_duplicate_slugs":[]},"computed_at":"2026-09-20T05:14:58.192Z","maintenance":{"label":"Active","score":82,"methodology":"github_public_v1","releases_90d":0,"days_since_push":11,"last_release_at":"2026-04-02T17:56:42Z","stars_delta_30d":-67,"open_issues_delta_30d":3},"security_summary":{"status":"findings","scanner":"osv@v1","low_count":2,"high_count":0,"last_scan_at":"2026-07-15T11:17:22.312Z","medium_count":0,"scan_profile":"deps","critical_count":0}},"capability_facts":{"mcp":{"source":"repo_scan","observed_at":"2026-09-20T05:14:59.221Z","server_manifest":false},"scan":{"source":"repo_scan","observed_at":"2026-09-20T05:14:59.221Z"},"has_cli":{"value":true,"source":"pyproject.toml:[project.scripts]","observed_at":"2026-09-20T05:14:59.221Z"},"languages":{"value":["python","javascript"],"source":"github.language+package.json+pyproject.toml","observed_at":"2026-09-20T05:14:59.221Z"},"license_spdx":{"value":"MIT","source":"github.license","observed_at":"2026-09-20T05:14:59.221Z"}},"decision_facts":{"hosting":null,"pricing":null,"requirements":null,"constraints":null,"when_to_use":["When optimizing the cost of running AI models by cascading less expensive models with more costly ones to balance quality.","If you require support for various AI model APIs such as Anthropic's Claude or HuggingFace, without needing complex integration setups."],"when_not_to_use":["In scenarios where strict control over the individual model's decision-making process is needed and cascading models might introduce complexity that negatively affects the desired outcome.","When working with a narrow range of AI use cases that do not benefit from cost optimization, as Cascadeflow's feature set provides less value."],"source":"enrich:decision_facts","observed_at":"2026-07-17T11:52:52.521Z"},"constraint_facets":null,"decision_summary":[{"label":"Adopt for","value":"Cascadeflow is an AI runtime optimized for cost and quality decisions within the agent loop, supporting multiple model APIs like Anthropic's Claude and HuggingFace."}]}}