Skip to content

DeepSeek API

DeepSeek is a hosted AI provider, not a coding-agent CLI — there is nothing to launch and no jsat connect deepseek/jsat deepseek command. It configures the AI backend JSAT's own LLM-backed tools use (crack, prompt_optimize, security_review, and so on), the same way anthropic/openai/gemini do.

DeepSeek's API is OpenAI-compatible, so JSAT reaches it through the same openai_compat backend used for Gemini and LM Studio — jsat ai use deepseek just presets the base URL and key environment variable for you.

Set API key

Get a key from platform.deepseek.com:

export DEEPSEEK_API_KEY=...

Activate

DeepSeek needs an explicit model — JSAT never guesses one:

jsat ai use deepseek --model deepseek-chat
# or the reasoning-focused model:
jsat ai use deepseek --model deepseek-reasoner

This sets the provider to openai_compat with base_url pointed at:

https://api.deepseek.com/v1

and api_key_env: DEEPSEEK_API_KEY, so DeepSeekProvider-style calls read the right key automatically — no manual api_key_env edit needed in .jsat/config.yaml.

# Apply globally (all projects on this machine):
jsat ai use deepseek --model deepseek-chat --global

Verify

jsat ai test "what is 2 + 2?"

Inside the shell

> switch deepseek <model>

Troubleshooting

  • A 404/model-not-found style error from jsat ai test means the model name is wrong — DeepSeek's own docs list the current model IDs; deepseek-chat and deepseek-reasoner are the two general-purpose options at time of writing.
  • If DEEPSEEK_API_KEY is unset, jsat ai use deepseek prints a warning at setup time; a missing/invalid key surfaces as an authentication error from jsat ai test.
  • This is unrelated to "DeepSeek Harness" (dsh), a separate open-source coding-agent CLI launchable via ollama launch dsh — JSAT does not currently wire MCP tools into that harness.