-
Notifications
You must be signed in to change notification settings - Fork 120
Expand file tree
/
Copy path.env.example
More file actions
88 lines (68 loc) · 3 KB
/
Copy path.env.example
File metadata and controls
88 lines (68 loc) · 3 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
# =============================================================
# InternAgent Environment Configuration
# Copy this file to .env and fill in the values you need.
# Not all keys are required — see comments for when each is used.
# =============================================================
# -------------------------------------------------------------
# Core LLM API (required)
# Used by all agents in the discovery and QA pipeline.
# Any OpenAI-compatible endpoint is supported via BASE_URL.
# -------------------------------------------------------------
OPENAI_API_KEY=
OPENAI_API_BASE_URL= # Required if using a third-party or self-hosted endpoint
# OPENAI_BASE_URL is an alias used by the sci_task scorer (ResearchClawBench).
# Set it to the same value as OPENAI_API_BASE_URL.
OPENAI_BASE_URL=
# OpenRouter (optional)
# Use this when models.default_provider is set to "openrouter" or when using
# config/openrouter_config.yaml.
OPENROUTER_API_KEY=
OPENROUTER_BASE_URL=https://openrouter.ai/api/v1
OPENROUTER_APP_NAME=InternAgent
OPENROUTER_SITE_URL=https://discovery.intern-ai.org.cn
# -------------------------------------------------------------
# Anthropic API (required for experiment backend)
# Used when running experiments with --exp_backend claudecode.
# -------------------------------------------------------------
ANTHROPIC_API_KEY=
# ANTHROPIC_BASE_URL= # Uncomment if using a proxy or custom endpoint
# -------------------------------------------------------------
# Alternative model providers (optional)
# Uncomment and fill in the provider you want to use,
# then set models.default_provider in config/default_config.yaml.
# -------------------------------------------------------------
# DeepSeek (R1 reasoning model)
# DS_API_KEY=
# DS_API_BASE_URL=
# InternLM / InternS1
# INS1_API_KEY=
# INS1_API_BASE_URL=
# Google Gemini (supported by the Deep Research agent)
# GEMINI_API_KEY=
# GEMINI_BASE_URL=
# -------------------------------------------------------------
# Deep Research — search tools (optional)
# Only the tools you enable in the DR config need keys.
# arxiv_search and openalex_search require no key.
# See docs/deep_research.md for details.
# -------------------------------------------------------------
# Volcengine web search (volc_search tool) — enterprise key required
# VOLC_SEARCH_API_KEY=
# Tavily URL extractor (url_processor tool)
# TAVILY_API_KEY=
# Serper web search (alternative to Volcengine)
# SERPER_API_KEY=
# Semantic Scholar literature search
# S2_API_KEY=
# Polite-access emails for free academic APIs (no key needed, just an email)
# OPENALEX_EMAIL=
# CROSSREF_EMAIL=
# -------------------------------------------------------------
# Model downloads (optional)
# Set HF_TOKEN to access private or gated HuggingFace models
# (e.g. embedding models for task memory).
# Set MODELSCOPE_SDK_TOKEN if downloading models from ModelScope
# (recommended for users in China).
# -------------------------------------------------------------
# HF_TOKEN=
# MODELSCOPE_SDK_TOKEN=