Deep Agents - Fireworks AI Docs

FireConnect routes LangChain Deep Agents Code (dcode) through Fireworks AI models. See the FireConnect overview for install and CLI basics.

Prerequisites

Deep Agents Code (dcode) installed
A Fireworks API key (fw_...) or a Fire Pass key (fpk_...)
The FireConnect CLI (see Install)

Azure routing not implemented yet for Deep Agents. fireconnect deepagents on always configures direct Fireworks, even when global config has --provider azure. See Microsoft Foundry in FireConnect.

Enable Fireworks routing

fireconnect login
fireconnect deepagents on
source ~/.zshrc   # keychain mode: load FIREWORKS_API_KEY for dcode

Or pass the key once:

fireconnect deepagents on --api-key fw_...

Restart dcode after on, model select, model reset, or off if it is already running.

fireconnect deepagents status

Default model

Deep Agents routes a single default model. The default is glm-fast-latest, written to config as fireworks:<model-id>.

What gets written

FireConnect edits ~/.deepagents/config.toml:

Sets [models].default to fireworks:<model> and configures [models.providers.fireworks] with the Fireworks OpenAI-compatible base URL (https://api.fireworks.ai/inference) and api_key_env = "FIREWORKS_API_KEY"
Stores your Fireworks API key in the FireConnect keychain (via login or deepagents on --api-key) and installs a shell hook that exports FIREWORKS_API_KEY for dcode
FireConnect does not write ~/.deepagents/.state/auth.json — use dcode’s /auth for credentials stored in that file

FireConnect snapshots config.toml under ~/.fireconnect/deepagents/ before the first change. Running fireconnect deepagents off restores it byte-for-byte.

Browsing and picking models

fireconnect deepagents model list              # browse serverless endpoints
fireconnect deepagents model select            # pick Deep Agents' default model
fireconnect deepagents model select --search glm

Fire Pass keys (fpk_...) show Fire Pass-supported routers only.

CLI reference

fireconnect deepagents on              # Enable Fireworks routing
fireconnect deepagents off             # Restore original config
fireconnect deepagents status          # Check current provider and model
fireconnect deepagents model list      # Browse serverless endpoints
fireconnect deepagents model select    # Pick a model interactively
fireconnect deepagents model reset     # Reset model to default
fireconnect deepagents help            # Show harness-specific help

Run fireconnect deepagents help for all options.

Switch models

fireconnect deepagents on --main glm-5p1

Turn off Fireworks routing

fireconnect deepagents off

This restores your previous config.toml from the backup in ~/.fireconnect/deepagents/.

Use a non-default config file

fireconnect deepagents on --config-path /path/to/config.toml

Source

FireConnect is open source: github.com/fw-ai/fireconnect

​Prerequisites

​Enable Fireworks routing

​Default model

​What gets written

​Browsing and picking models

​CLI reference

​Switch models

​Turn off Fireworks routing

​Use a non-default config file

​Source

Prerequisites

Enable Fireworks routing

Default model

What gets written

Browsing and picking models

CLI reference

Switch models

Turn off Fireworks routing

Use a non-default config file

Source