Repository navigation
Expand file tree
/
Copy path.env.example
More file actions
54 lines (48 loc) · 3.57 KB
/
Copy path.env.example
File metadata and controls
54 lines (48 loc) · 3.57 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
# Copy to .env and fill in ONE provider's key - Writing Agent has no blessed default; you choose
# the host with the first-run wizard, `/provider <id>`, the `provider:` setting, or
# WRITINGAGENT_PROVIDER. Set the key for whichever host you pick (only that one is needed).
#
# Optionally pin the host up front:
# WRITINGAGENT_PROVIDER=openrouter # openrouter | openai | anthropic | deepseek | google | ...
# ── Aggregators (one key, many vendors) ──────────────────────────────────────
OPENROUTER_API_KEY=sk-or-... # openrouter.ai/keys - 300+ models, reports real USD cost
# TOGETHER_API_KEY=...
# FIREWORKS_API_KEY=...
# DEEPINFRA_API_KEY=...
# ── First-party APIs (pick your vendor) ──────────────────────────────────────
# OPENAI_API_KEY=sk-... # platform.openai.com/api-keys (gpt-*)
# ANTHROPIC_API_KEY=sk-ant-... # console.anthropic.com/settings/keys (claude-*)
# DEEPSEEK_API_KEY=... # platform.deepseek.com/api_keys
# GEMINI_API_KEY=... # aistudio.google.com/app/apikey (gemini-*)
# XAI_API_KEY=... # x.ai (grok-*)
# GROQ_API_KEY=... # console.groq.com/keys
# CEREBRAS_API_KEY=...
# SAMBANOVA_API_KEY=...
# PERPLEXITY_API_KEY=... # sonar (answers with live citations)
# MISTRAL_API_KEY=...
# MOONSHOT_API_KEY=... # Kimi
# DASHSCOPE_API_KEY=... # Qwen / Alibaba
# ZHIPU_API_KEY=... # GLM
# NVIDIA_API_KEY=... # NIM
# ── Enterprise clouds (via an OpenAI-compatible gateway - see providers.py notes) ──
# AWS_BEDROCK_BASE_URL=http://localhost:8000/v1 # LiteLLM / bedrock-access-gateway
# AWS_BEDROCK_API_KEY=... # the gateway's key
# AZURE_OPENAI_BASE_URL=https://<proxy>/v1
# AZURE_OPENAI_API_KEY=...
# ── Local servers (no key needed) ────────────────────────────────────────────
# provider: ollama (run `ollama serve`) · provider: lmstudio (start the LM Studio server)
# ── Any other OpenAI-compatible server ───────────────────────────────────────
# WRITINGAGENT_BASE_URL=https://your-gateway/v1
# WRITINGAGENT_API_KEY=...
# ── Optional: web search backends for the researcher ─────────────────────────
# DuckDuckGo is the keyless default. Set `search_provider` (in settings.yaml or the
# dashboard) to one of the below and add its key; a missing key degrades to DuckDuckGo.
# FIRECRAWL_API_KEY=fc-... # search_provider: firecrawl (also scrapes pages in deep-research)
# TAVILY_API_KEY=tvly-... # search_provider: tavily (built for LLM agents)
# BRAVE_API_KEY=... # search_provider: brave (brave.com/search/api)
# SERPAPI_API_KEY=... # search_provider: serpapi (Google results via serpapi.com)
# EXA_API_KEY=... # search_provider: exa (neural/semantic search)
# PARALLEL_API_KEY=... # search_provider: parallel (parallel.ai web search, beta)
#
# Image generation reuses your model host's key (no separate key). Set image_source: generate
# and image_model: <slug> (e.g. gpt-image-1) - and optionally image_provider for a different host.