External agent
esterno An agent installed on this computer.
- Protocol
- Agent Client Protocol
- Key
- None
- Timeout
- 180 s by default
The Archive · CLI reference
Providers
The 29 providers of TALOS CLI: protocol, key variables, endpoint and sign-in.
Reference for TALOS CLI 0.3.3
TALOS CLI 0.5.2 is out; this page still describes 0.3.3 and updates with the next build.
Set a key with talos provider set-key <id>, check it with talos provider test <id>, change an endpoint with talos provider configure <id>. See talos provider.
OpenAI
openai OpenAI models, direct API (Responses).
- Protocol
- OpenAI Responses
- Key
- API key (Bearer)
- Key variables
OPENAI_API_KEY- Endpoint
https://api.openai.com/v1· editable- Timeout
- 60 s by default
DeepSeek
deepseek DeepSeek models, direct API (OpenAI-compatible).
- Protocol
- OpenAI Chat Completions
- Key
- API key (Bearer)
- Key variables
DEEPSEEK_API_KEY- Endpoint
https://api.deepseek.com· editable- Timeout
- 60 s by default
Kimi
kimi Kimi models, direct international connection.
- Protocol
- OpenAI Chat Completions
- Key
- API key (Bearer)
- Key variables
MOONSHOT_API_KEYKIMI_API_KEY- Endpoint
https://api.moonshot.ai/v1· editable- Endpoint variables
MOONSHOT_BASE_URL- Timeout
- 60 s by default
MiniMax
minimax MiniMax models, direct international connection.
- Protocol
- OpenAI Chat Completions
- Key
- API key (Bearer)
- Key variables
MINIMAX_API_KEY- Endpoint
https://api.minimax.io/v1· editable- Endpoint variables
MINIMAX_BASE_URL- Timeout
- 60 s by default
Qwen
qwen Qwen models, direct international connection.
- Protocol
- OpenAI Chat Completions
- Key
- API key (Bearer)
- Key variables
DASHSCOPE_API_KEY- Endpoint
https://dashscope-intl.aliyuncs.com/compatible-mode/v1· editable- Endpoint variables
DASHSCOPE_BASE_URL- Timeout
- 60 s by default
Z.AI
zai GLM models through the Z.AI direct API.
- Protocol
- OpenAI Chat Completions
- Key
- API key (Bearer)
- Key variables
GLM_API_KEYZAI_API_KEYZ_AI_API_KEY- Endpoint
https://api.z.ai/api/paas/v4· editable- Timeout
- 60 s by default
- Keys
- z.ai
Anthropic
anthropic Anthropic models, direct API (Messages).
- Protocol
- Anthropic Messages
- Key
- API key (header)
- Key variables
ANTHROPIC_API_KEY- Endpoint
https://api.anthropic.com/v1- Timeout
- 60 s by default
Google Gemini
gemini Gemini models, direct Google AI Studio API.
- Protocol
- Gemini
- Key
- API key (query)
- Key variables
GEMINI_API_KEYGOOGLE_API_KEY- Endpoint
https://generativelanguage.googleapis.com/v1beta- Timeout
- 60 s by default
OpenRouter
openrouter Aggregator: hundreds of models behind a single key.
- Protocol
- OpenAI Chat Completions
- Key
- API key (Bearer)
- Key variables
OPENROUTER_API_KEY- Endpoint
https://openrouter.ai/api/v1· editable- Sign-in
- in the browser:
talos provider login openrouter - Timeout
- 60 s by default
- Keys
- openrouter.ai
Ollama Local
ollama Local Ollama engine on this computer.
- Protocol
- OpenAI Chat Completions
- Key
- API key (Bearer) · optional
- Key variables
OLLAMA_API_KEY- Endpoint
http://127.0.0.1:11434· editable- Endpoint variables
OLLAMA_BASE_URL- Timeout
- 60 s by default
- Keys
- ollama.com
LM Studio
lmstudio Local LM Studio engine on this computer.
- Protocol
- OpenAI Chat Completions
- Key
- API key (Bearer) · optional
- Key variables
LMSTUDIO_API_KEYLM_API_KEY- Endpoint
http://127.0.0.1:1234· editable- Endpoint variables
LMSTUDIO_BASE_URLLM_BASE_URL- Timeout
- 60 s by default
- Keys
- lmstudio.ai
Groq
groq Models through Groq.
- Protocol
- OpenAI Chat Completions
- Key
- API key (Bearer)
- Key variables
GROQ_API_KEY- Endpoint
https://api.groq.com/openai/v1· editable- Timeout
- 60 s by default
- Keys
- console.groq.com
Cerebras
cerebras Models through Cerebras.
- Protocol
- OpenAI Chat Completions
- Key
- API key (Bearer)
- Key variables
CEREBRAS_API_KEY- Endpoint
https://api.cerebras.ai/v1· editable- Timeout
- 60 s by default
Mistral
mistral Models through Mistral.
- Protocol
- OpenAI Chat Completions
- Key
- API key (Bearer)
- Key variables
MISTRAL_API_KEY- Endpoint
https://api.mistral.ai/v1· editable- Timeout
- 60 s by default
Together
together Models through Together.
- Protocol
- OpenAI Chat Completions
- Key
- API key (Bearer)
- Key variables
TOGETHER_API_KEY- Endpoint
https://api.together.ai/v1· editable- Timeout
- 60 s by default
- Keys
- api.together.ai
Fireworks
fireworks Models through Fireworks.
- Protocol
- OpenAI Chat Completions
- Key
- API key (Bearer)
- Key variables
FIREWORKS_API_KEY- Endpoint
https://api.fireworks.ai/inference/v1· editable- Timeout
- 60 s by default
- Keys
- app.fireworks.ai
DeepInfra
deepinfra Models through DeepInfra.
- Protocol
- OpenAI Chat Completions
- Key
- API key (Bearer)
- Key variables
DEEPINFRA_API_KEYDEEPINFRA_TOKEN- Endpoint
https://api.deepinfra.com/v1/openai· editable- Timeout
- 60 s by default
- Keys
- deepinfra.com
Novita
novita Models through Novita.
- Protocol
- OpenAI Chat Completions
- Key
- API key (Bearer)
- Key variables
NOVITA_API_KEY- Endpoint
https://api.novita.ai/openai/v1· editable- Timeout
- 60 s by default
- Keys
- novita.ai
Nebius
nebius Models through Nebius.
- Protocol
- OpenAI Chat Completions
- Key
- API key (Bearer)
- Key variables
NEBIUS_API_KEY- Endpoint
https://api.tokenfactory.nebius.com/v1· editable- Timeout
- 60 s by default
xAI
xai Models through xAI.
- Protocol
- OpenAI Chat Completions
- Key
- API key (Bearer)
- Key variables
XAI_API_KEY- Endpoint
https://api.x.ai/v1· editable- Timeout
- 60 s by default
- Keys
- console.x.ai
Ollama Cloud
ollama-cloud Models through Ollama Cloud.
- Protocol
- OpenAI Chat Completions
- Key
- API key (Bearer)
- Key variables
OLLAMA_API_KEY- Endpoint
https://ollama.com/v1· editable- Timeout
- 60 s by default
- Keys
- ollama.com
Hugging Face
huggingface Hugging Face model inference, catalog and downloads.
- Protocol
- OpenAI Chat Completions
- Key
- API key (Bearer)
- Key variables
HF_TOKENHUGGINGFACE_HUB_TOKEN- Endpoint
https://router.huggingface.co/v1· editable- Timeout
- 60 s by default
- Keys
- huggingface.co
Z.AI (Anthropic endpoint)
zai-anthropic GLM models through the Z.AI coding plan endpoint.
- Protocol
- Anthropic Messages
- Key
- API key (Bearer)
- Key variables
ZAI_ANTHROPIC_API_KEY- Endpoint
https://api.z.ai/api/anthropic/v1· editable- Endpoint variables
ZAI_ANTHROPIC_BASE_URL- Timeout
- 60 s by default
- Keys
- z.ai
MiniMax (Anthropic endpoint)
minimax-anthropic MiniMax models through the Anthropic-compatible endpoint (international access).
- Protocol
- Anthropic Messages
- Key
- API key (header)
- Key variables
MINIMAX_API_KEY- Endpoint
https://api.minimax.io/anthropic/v1· editable- Endpoint variables
MINIMAX_BASE_URL- Timeout
- 60 s by default
Azure AI Foundry
azure Models from your own Azure resource.
- Protocol
- OpenAI Chat Completions
- Key
- API key (header)
- Key variables
AZURE_OPENAI_API_KEYAZURE_FOUNDRY_API_KEY- Endpoint
- yours, set with
talos provider configure - Endpoint variables
AZURE_OPENAI_ENDPOINTAZURE_FOUNDRY_BASE_URL- Timeout
- 60 s by default
- Keys
- ai.azure.com
Amazon Bedrock
bedrock Models available in the chosen region.
- Protocol
- OpenAI Chat Completions
- Key
- API key (Bearer)
- Key variables
AWS_BEARER_TOKEN_BEDROCK- Endpoint
- yours, set with
talos provider configure - Endpoint variables
BEDROCK_OPENAI_BASE_URL- Timeout
- 60 s by default
Google Vertex AI
vertex Models of the chosen project and region.
- Protocol
- OpenAI Chat Completions
- Key
- API key (Bearer)
- Key variables
VERTEX_ACCESS_TOKEN- Endpoint
- yours, set with
talos provider configure - Endpoint variables
VERTEX_OPENAI_BASE_URL- Timeout
- 60 s by default
Local engine (llama.cpp)
local The llama-server supervisor of this computer.
- Protocol
- llama.cpp server
- Key
- None (local)
- Timeout
- 60 s by default