Set up Claude Code

01 / PREPARE

Install the agent and prepare access

1. Install the agent using its official guide. Check its version before merging the examples below.
2. Purchase your Multi-Model plan, then open Billing & API keys to copy your active key. OpenAI base: https://multi-api.savemytoken.com/v1. Claude Code base: https://multi-api.savemytoken.com. DeepSeek credentials are a separate scope.
3. Back up the file you will edit. Merge fields into existing objects and tables; preserve unrelated providers, tools and permission settings. Keep credentials in private local settings.
4. Image input is supported by seven routes; GLM-5.3 is text-only. Send JPEG, PNG, GIF or WebP images as Base64 or public HTTP(S) image URLs. The gateway accepts up to 20 images, 10 MiB per inline image and 20 MiB total inline image data per request.
5. Each image temporarily reserves 16,384 input tokens at the selected model's points rate. After completion, the reservation is replaced by the supplier's actual input and output token usage, including image tokens. Base64 characters are not billed as text tokens.

Ready when: the agent opens, and you know which plan, process and configuration directory you will use.

02 / CONFIGURE

Save the connection and model catalog

Create the parent directories if needed. Download or copy each file, replace the example endpoint and local placeholders, then merge it into the stated location. Downloads and the examples below contain the same configuration.

{
  "model": "deepseek-flash",
  "modelPicker": {
    "replaceBuiltInOptions": true,
    "options": [
      {
        "model": "deepseek-flash",
        "label": "DeepSeek Flash",
        "description": "deepseek-flash"
      },
      {
        "model": "glm-5.3-flash",
        "label": "GLM-5.3-Flash",
        "description": "glm-5.3-flash"
      },
      {
        "model": "mimo-flash",
        "label": "MiMo V2.6 Flash",
        "description": "mimo-flash"
      },
      {
        "model": "mimo-pro",
        "label": "MiMo V2.6 Pro",
        "description": "mimo-pro"
      },
      {
        "model": "glm-5.3",
        "label": "GLM-5.3",
        "description": "glm-5.3"
      },
      {
        "model": "kimi-k3",
        "label": "Kimi-K3",
        "description": "kimi-k3"
      },
      {
        "model": "qwen3.8-max",
        "label": "Qwen3.8 Max",
        "description": "qwen3.8-max"
      },
      {
        "model": "minimax-m3",
        "label": "MiniMax M3",
        "description": "minimax-m3"
      }
    ]
  },
  "env": {
    "ANTHROPIC_BASE_URL": "https://multi-api.savemytoken.com",
    "ANTHROPIC_AUTH_TOKEN": "REPLACE_WITH_YOUR_DELIVERED_KEY",
    "ANTHROPIC_DEFAULT_OPUS_MODEL": "deepseek-flash",
    "ANTHROPIC_DEFAULT_SONNET_MODEL": "deepseek-flash",
    "ANTHROPIC_DEFAULT_HAIKU_MODEL": "deepseek-flash"
  }
}

Ready when: the files parse successfully and the selected profile points to the delivered URL, key reference and exact model catalog.

03 / START

Start the agent and choose your model

Open a terminal in a trusted project. Complete the file setup before running these commands.

mkdir -p "$HOME/.config/savemytoken/multi-model/claude"
# Save settings.json above, then protect it:
chmod 600 "$HOME/.config/savemytoken/multi-model/claude/settings.json"
CLAUDE_CONFIG_DIR="$HOME/.config/savemytoken/multi-model/claude" claude

Open /model for the named catalog, or type /model followed by an exact model ID. Opus, Sonnet and Haiku are fixed shortcuts in this configuration; they all start on DeepSeek Flash and do not follow later model switches. These aliases do not supply Anthropic models.

Check: the provider is the one you configured and the selected model is deepseek-flash. A visible model list alone does not prove inference works.

04 / REASONING

Choose and verify reasoning effort

Use /effort to inspect and change intensity. Current versions can save a per-model default; use the session-only selector or --effort for a temporary override. Max is session-only.

Remove a fixed CLAUDE_CODE_EFFORT_LEVEL if you want interactive control. Review modelSettings, effortLevel and disabled thinking settings when the choice does not take effect.

The client may offer more levels than your backend supports. Use the capability table below. /effort auto clears a manual intensity; it is not MiniMax's adaptive thinking mode and does not guarantee that the client omits effort.

/status
/model deepseek-flash
/effort status
/effort low
/effort high

On DeepSeek Flash, choose Low, then High, then Low again. Inspect the active setting each time before sending the next message. A longer answer is not evidence that effort changed.

REASONING CONTROLS

Choose effort for the actual model

Use these exact gateway IDs. All eight models draw from the same points allowance. Image input support is listed per model below. Output is text or tool calls; audio and video input and media generation are not included.

05 / DESKTOP

Claude Desktop · Code tab

Claude Desktop uses its own inference settings, not this CLI settings.json. Use Help → Troubleshooting → Enable Developer Mode, restart, then Developer → Configure Third-Party Inference when supported by your app.

Choose Gateway and use the separately delivered Desktop compatibility URL with Bearer authentication. Enable model discovery and test it. Use the IDs and backend labels returned by that endpoint; do not invent Claude-style aliases for gateway models.

Save and restart, open Code, then start a local task. Check the actual backend and the effort control for that alias. If a required effort is unavailable, use the CLI profile. SaveMyToken has not yet published a Desktop compatibility endpoint or alias mapping.