> ## Documentation Index
> Fetch the complete documentation index at: https://docs.fireworks.ai/llms.txt
> Use this file to discover all available pages before exploring further.

# Claude Code

> Use Fireworks AI models in Claude Code with the FireConnect CLI

[FireConnect](https://github.com/fw-ai/fireconnect) routes [Claude Code](https://claude.ai/code) through Fireworks AI models. See the [FireConnect overview](/ecosystem/fireconnect/overview) for install and CLI basics.

## Prerequisites

* [Claude Code](https://claude.ai/code) installed
* A [Fireworks API key](https://app.fireworks.ai/settings/users/api-keys) (`fw_...`) or a [Fire Pass](/firepass) key (`fpk_...`)
* The FireConnect CLI (see [Install](/ecosystem/fireconnect/overview#install))

<Note>
  **Azure routing not implemented yet for Claude Code.** `fireconnect claude on` always configures direct Fireworks, even when global config has `--provider azure` or you pass `--azure`. See [Microsoft Foundry in FireConnect](/ecosystem/fireconnect/microsoft-foundry#supported-harnesses).
</Note>

## Enable Fireworks routing

```bash theme={null}
fireconnect login
fireconnect claude on
```

Or pass the key once:

```bash theme={null}
fireconnect claude on --api-key fw_...
```

Restart Claude Code after enabling, then test with:

```text theme={null}
hi
```

After `fireconnect claude on`, `model select`, or `model reset`, settings are updated immediately. To use a new model in the same session, run `/model` in Claude Code, start a new session, or `/exit` and resume with `claude --resume <id>`.

## Using Fire Pass

Use your `fpk_...` key during `login` or with `--api-key`:

```bash theme={null}
fireconnect claude on --api-key fpk_...
```

FireConnect detects Fire Pass keys and routes all model aliases to `glm-fast-latest`.

## Default model mapping

| Alias    | Standard key (`fw_...`) | Fire Pass key (`fpk_...`) |
| -------- | ----------------------- | ------------------------- |
| main     | `glm-fast-latest`       | `glm-fast-latest`         |
| opus     | `glm-fast-latest`       | `glm-fast-latest`         |
| fable    | `glm-fast-latest`       | `glm-fast-latest`         |
| sonnet   | `glm-5p1`               | `glm-fast-latest`         |
| haiku    | `deepseek-v4-flash`     | `glm-fast-latest`         |
| subagent | `deepseek-v4-flash`     | `glm-fast-latest`         |

Short model IDs like `glm-fast-latest` are expanded to full Fireworks paths (for example, `accounts/fireworks/routers/glm-fast-latest[1m]`). FireConnect appends the `[1m]` suffix on `main`, `opus`, and `fable` so Claude Code enables 1M context. The `subagent` slot is written without `[1m]` because Claude Code forwards that value verbatim to the provider API.

## What gets written

FireConnect writes these settings to `~/.claude/settings.json` (no literal API key in the file — the key lives in the OS keychain and is fetched at runtime via `apiKeyHelper`):

```json theme={null}
{
  "apiKeyHelper": "/path/to/fireconnect key export",
  "env": {
    "ANTHROPIC_BASE_URL": "https://api.fireworks.ai/inference",
    "ANTHROPIC_MODEL": "accounts/fireworks/routers/glm-fast-latest[1m]",
    "ANTHROPIC_DEFAULT_OPUS_MODEL": "accounts/fireworks/routers/glm-fast-latest[1m]",
    "ANTHROPIC_DEFAULT_SONNET_MODEL": "accounts/fireworks/models/glm-5p1",
    "ANTHROPIC_DEFAULT_HAIKU_MODEL": "accounts/fireworks/models/deepseek-v4-flash",
    "ANTHROPIC_DEFAULT_FABLE_MODEL": "accounts/fireworks/routers/glm-fast-latest[1m]",
    "CLAUDE_CODE_SUBAGENT_MODEL": "accounts/fireworks/models/deepseek-v4-flash"
  }
}
```

Claude Code calls `apiKeyHelper` at runtime to fetch the key from the OS keychain. FireConnect saves a backup of your previous provider settings to `~/.fireconnect/claude/` so `fireconnect claude off` can restore them.

## Browsing and picking models

```bash theme={null}
fireconnect claude model list              # browse callable serverless endpoints
fireconnect claude model select            # pick a model for Claude Code
fireconnect claude model select --slot sonnet   # update one alias
```

### `fireconnect claude model list`

Lists serverless models from the Fireworks API (`supports_serverless=true`) and merges known public platform routers (`glm-latest`, `glm-fast-latest`, `glm-5p2-fast`, `kimi-fast-latest`, `kimi-latest`, `kimi-k2p6-turbo`, and `kimi-k2p7-code-fast`). Every row is tagged `serverless`.

```bash theme={null}
fireconnect claude model list
fireconnect claude model list --search glm
fireconnect claude model list --json
```

Resolves the key from `--api-key`, the OS keychain, `~/.fireconnect/config.json`, or `FIREWORKS_API_KEY`. Fire Pass keys (`fpk_...`) show Fire Pass-supported routers only.

### `fireconnect claude model select`

Interactive picker. Requires a terminal and Fireworks to be enabled.

```bash theme={null}
fireconnect claude model select
fireconnect claude model select --slot sonnet
fireconnect claude model select --slot sonnet --search glm
```

### `fireconnect claude status` vs `fireconnect claude model list`

| Command                         | Shows                                                                                          |
| ------------------------------- | ---------------------------------------------------------------------------------------------- |
| `fireconnect claude status`     | Your current provider, auth, configured alias mapping, and Fireworks serverless rates per slot |
| `fireconnect claude model list` | Available serverless endpoints from the Fireworks API, with IN / OUT pricing where known       |

## Claude Code pricing estimates

Claude Code's `/model` picker and session cost estimates use **Anthropic list prices**, not Fireworks serverless rates. For example, the default `glm-fast-latest` mapping may show Opus-tier estimates around **$5 / $25 per Mtok** while Fireworks bills at model-specific serverless rates (often much lower).

FireConnect cannot override Claude Code's price column. Use `fireconnect claude status` and `fireconnect claude model list` for Fireworks rates, and check the [billing dashboard](https://app.fireworks.ai/account/billing) for actual spend.

FireConnect also writes `ANTHROPIC_CUSTOM_MODEL_OPTION_DESCRIPTION` with Fireworks rates for your **main** model — that text appears as the subtitle on the custom picker entry at the bottom of `/model`, not in the price column.

## CLI reference

```bash theme={null}
fireconnect claude on         # Route Claude Code through Fireworks
fireconnect claude off        # Restore your previous provider
fireconnect claude status     # Show the current provider and model mapping
fireconnect claude usage      # Estimate usage cost from a session log
fireconnect claude model list # Browse serverless models
fireconnect claude model select   # Pick a model interactively
fireconnect claude model reset    # Reset model aliases to defaults
fireconnect claude help       # Show harness-specific help
```

Run `fireconnect claude help` for all options.

### Switch models

```bash theme={null}
fireconnect claude on --main glm-fast-latest --sonnet glm-5p1 --haiku deepseek-v4-flash --subagent deepseek-v4-flash
```

Or pick interactively:

```bash theme={null}
fireconnect claude model select --slot opus
```

### Turn off Fireworks routing

```bash theme={null}
fireconnect claude off
```

This restores your previous `~/.claude/settings.json` from the backup saved in `~/.fireconnect/claude/`.

## Uninstall

To remove FireConnect from your machine entirely (all harnesses):

```bash theme={null}
fireconnect uninstall
```

## Source

FireConnect is open source: [github.com/fw-ai/fireconnect](https://github.com/fw-ai/fireconnect)