# DeerFlow Open agent runtime for research, coding and multi-step tasks on your own infrastructure. Canonical profile: https://ki-agentenkompass.de/en/agents/deerflow/ Editor: [Michael Meierhoff](https://ki-agentenkompass.de/en/about/) Sources reviewed: 2026-10-26 Basis: provider documentation. Examples are editorial suggestions, not hands-on product tests. ## Overview Available offering; access methods and prerequisites are described separately in this profile. - Provider: ByteDance - Product family: DeerFlow - Profile type: Agent platform - Environment: Own computer or server - Connections: Subagents, Sandbox, Memory, Skills, Research and code - Setup: Requires your own installation, model access and configured tools. - Pricing model: MIT-licensed source code; model usage and infrastructure are funded separately. ## Key features ### Subagents Delegate complex tasks to specialized subagents. Source: https://github.com/bytedance/deer-flow ### Sandbox Run tools in configured execution environments. Source: https://github.com/bytedance/deer-flow ### Memory Store context across sessions. Source: https://github.com/bytedance/deer-flow ### Skills Extend workflows through skills. Source: https://github.com/bytedance/deer-flow ### Research and code Combine research and coding with tools. Source: https://github.com/bytedance/deer-flow ## Example workflows ### Subagents - Your input: Delegate complex tasks to specialized subagents. - Possible result: A work product for your review. - What to check: Review results and write actions before accepting them. ### Sandbox - Your input: Run tools in configured execution environments. - Possible result: A work product for your review. - What to check: Review results and write actions before accepting them. ## Requirements and limits ### Setup Requirement: Requires your own installation, model access and configured tools. Source: https://github.com/bytedance/deer-flow ### Usage costs Limitation: MIT-licensed source code; model usage and infrastructure are funded separately. Source: https://github.com/bytedance/deer-flow ### Output quality Not yet confirmed: Review results and write actions before accepting them. Source: https://github.com/bytedance/deer-flow ## A good fit if … - you want to run an agent runtime yourself with Docker or locally - you want to start tasks via Slack, Telegram or Feishu - you want to trace runs with LangSmith or Langfuse ## Less suitable if … - you want a hosted service with support and a service agreement - you require documented approval before every write action Editorial assessment based on provider documentation. Not a product test. ## Usable AI models Open choice DeerFlow ships with no fixed model: you enter provider and key yourself, in the config or through presets in the web UI. This works for OpenAI, Anthropic, Gemini, DeepSeek, Qwen, Kimi, MiniMax, GLM and Doubao, for OpenRouter and any OpenAI-compatible endpoint. Ollama, llama.cpp and vLLM run locally, and the Codex and Claude Code CLIs can also be connected. - Model families named: OpenAI (GPT), Anthropic (Claude), Google (Gemini), DeepSeek, Alibaba (Qwen), Moonshot (Kimi), MiniMax, Z.ai (GLM), In-house model · Other models can be connected freely - Model choice: You choose - Your own API key: Yes - Router or custom endpoint: Yes - Local models: Yes Documented · Sources accessed on: 2026-11-05 - [DeerFlow: README](https://raw.githubusercontent.com/bytedance/deer-flow/main/README.md) - [DeerFlow: config.example.yaml](https://raw.githubusercontent.com/bytedance/deer-flow/main/config.example.yaml) ## Europe & regional availability Consider installation and provider access separately. A complete regional feature matrix is not published. Reviewed scope: Product access in Germany and the EU Reviewed on: 2026-10-26 ### Access and connected services With requirements Requires your own installation, model access and configured tools. Region covered: Germany / EU Documented reason: The reviewed sources document no specific restriction caused by the AI Act, DMA or GDPR. Model and service access depend on country and contract. Official source: https://github.com/bytedance/deer-flow ## Accounts & permissions Based on provider documentation. Permissions depend on the account, integration and organization policies. Review the permissions actually requested before use. Configuration covered: Self-hosted DeerFlow from the open-source repository Sources reviewed: 2026-11-05 ### Account used Documented: You enter models in config.yaml and refer to environment variables for keys. Credentials stored in the model catalogue are encrypted with a key from DEER_FLOW_CREDENTIALS_KEY. Local logins use a JWT secret, and failed logins are throttled per IP. Evidence: https://github.com/bytedance/deer-flow/blob/main/backend/docs/CONFIGURATION.md Evidence: https://github.com/bytedance/deer-flow/blob/main/Install.md ### Admin approval Documented with conditions: Some functions are restricted to administrators: enabling and disabling skills, plugin diagnostics and managing the shared model catalogue. The pages mention no approval by a provider. Evidence: https://github.com/bytedance/deer-flow/blob/main/backend/README.md Evidence: https://github.com/bytedance/deer-flow/blob/main/backend/docs/CONFIGURATION.md ### Data access & team sharing Documented with conditions: The local sandbox provider has the shell disabled by default; the Docker provider offers isolated shell access. According to the configuration documentation local mode runs directly on the host and is not a strong isolation boundary. Message channels are Feishu, Slack and Telegram. Evidence: https://github.com/bytedance/deer-flow/blob/main/backend/README.md Evidence: https://github.com/bytedance/deer-flow/blob/main/backend/docs/CONFIGURATION.md ## Prices and included usage Pricing checked: 2026-10-26 MIT-licensed source code; model usage and infrastructure are funded separately. MIT-licensed source code; model usage and infrastructure are funded separately. Pricing details: https://github.com/bytedance/deer-flow ## Operation & oversight Based on public provider documentation. Assessments are editorial inferences from the stated properties. Actual time requirements and reliability have not been measured. Deployment covered: Self-hosted instance with gateway, Docker or local operation Sources reviewed: 2026-11-05 ### Files and working state Documented with conditions: Application tables for runs, threads and run events live in SQLite or Postgres. The memory feature extracts facts and preferences from conversations into a JSON file and injects top facts into prompts. The sandbox is per thread. Evidence: https://github.com/bytedance/deer-flow/blob/main/backend/README.md ### Sign-ins Open in reviewed sources: The configuration page names no validity period for sessions or service connections. In the managed model catalogue a missing or wrong key fails closed; other stored values are then treated as missing. Evidence: https://github.com/bytedance/deer-flow/blob/main/backend/docs/CONFIGURATION.md ### Starting further runs Documented with conditions: The backend README names message channels (Feishu, Slack, Telegram) for access. The configuration documentation mentions a scheduler that sends notices to connected channels; how schedules are created is not stated there. Evidence: https://github.com/bytedance/deer-flow/blob/main/backend/README.md Evidence: https://github.com/bytedance/deer-flow/blob/main/backend/docs/CONFIGURATION.md ### Human intervention Documented with conditions: An ask_clarification tool lets the agent ask questions, and middleware then interrupts execution. The backend README names no general approval before write actions. Evidence: https://github.com/bytedance/deer-flow/blob/main/backend/README.md ### Detecting errors Documented with conditions: LangSmith and Langfuse can be used for tracing; if a provider is explicitly enabled with invalid credentials, DeerFlow raises an error instead of silently disabling tracing. An endpoint returns debug and audit events per run. Evidence: https://github.com/bytedance/deer-flow/blob/main/backend/README.md ### Resuming after failures Documented with conditions: After a sandbox container ends unexpectedly, the Docker provider drops the dead entry so a thread can get a fresh sandbox. Durable batch tasks keep their lease during initialisation, execution and acceptance checks. The README does not describe resuming interrupted runs. Evidence: https://github.com/bytedance/deer-flow/blob/main/backend/README.md ### Editorial assessment You run the database, keys and tracing yourself. The backend README names no general approval before write actions, so restrict tool permissions and review results yourself. ## Memory and conversation continuity Scope: Public main branch; pluggable memory backends Sources accessed: 2026-11-04 Automatically extracted facts with scope checks, manageable entries and backend selection. ### Storage and scope Documented: The backend README describes structured user context, history and scored facts in JSON. Newer backends are pluggable; storage depends on the configured manager. - [DeerFlow backend README](https://github.com/bytedance/deer-flow/blob/main/backend/README.md) - [Memory API implementation](https://github.com/bytedance/deer-flow/blob/main/backend/app/gateway/routers/memory.py) - [Memory backends](https://github.com/bytedance/deer-flow/blob/main/backend/packages/harness/deerflow/agents/memory/backends/README.md) ### Selection and timing Documented: Middleware automatically extracts durable user facts and batches updates with a configurable delay. Scope and confidence checks limit promotion. - [DeerFlow backend README](https://github.com/bytedance/deer-flow/blob/main/backend/README.md) - [Memory API implementation](https://github.com/bytedance/deer-flow/blob/main/backend/app/gateway/routers/memory.py) - [Memory backends](https://github.com/bytedance/deer-flow/blob/main/backend/packages/harness/deerflow/agents/memory/backends/README.md) ### Inspect, edit and delete Documented: The API reads, creates, updates and deletes facts and the full store. Support for individual management operations depends on the memory backend. - [DeerFlow backend README](https://github.com/bytedance/deer-flow/blob/main/backend/README.md) - [Memory API implementation](https://github.com/bytedance/deer-flow/blob/main/backend/app/gateway/routers/memory.py) - [Memory backends](https://github.com/bytedance/deer-flow/blob/main/backend/packages/harness/deerflow/agents/memory/backends/README.md) ### Agent and channel boundaries Documented: API access resolves a user identity; automatic global promotion requires durable user facts. Task- or project-specific facts are excluded. - [DeerFlow backend README](https://github.com/bytedance/deer-flow/blob/main/backend/README.md) - [Memory API implementation](https://github.com/bytedance/deer-flow/blob/main/backend/app/gateway/routers/memory.py) - [Memory backends](https://github.com/bytedance/deer-flow/blob/main/backend/packages/harness/deerflow/agents/memory/backends/README.md) ### Retrieval Documented: Documented behavior injects important facts and context into prompts. Pluggable backends may provide search; the selected operating mode determines retrieval. - [DeerFlow backend README](https://github.com/bytedance/deer-flow/blob/main/backend/README.md) - [Memory API implementation](https://github.com/bytedance/deer-flow/blob/main/backend/app/gateway/routers/memory.py) - [Memory backends](https://github.com/bytedance/deer-flow/blob/main/backend/packages/harness/deerflow/agents/memory/backends/README.md) ### Session reset and compaction Partly documented: Memory spans conversations; summarization separately configures context summaries. Reset and retention depend on the chosen backend. - [DeerFlow backend README](https://github.com/bytedance/deer-flow/blob/main/backend/README.md) - [Memory API implementation](https://github.com/bytedance/deer-flow/blob/main/backend/app/gateway/routers/memory.py) - [Memory backends](https://github.com/bytedance/deer-flow/blob/main/backend/packages/harness/deerflow/agents/memory/backends/README.md) ## Context Review results and write actions before accepting them. Official sources reviewed. Example tasks are editorial suggestions; hands-on testing is outstanding. ## Open points Requirements and limits, not yet confirmed: - Output quality ## Sources - https://github.com/bytedance/deer-flow ## Provider Open agent runtime for research, coding and multi-step tasks on your own infrastructure. - Operator: ByteDance - Organization: Provider - Website: https://github.com/bytedance/deer-flow - Reviewed: 2026-10-26 - Provider source: https://github.com/bytedance/deer-flow [Sources and method](https://ki-agentenkompass.de/en/sources/)