TokenmaxerOpencode
Point Opencode at a Tokenmaxer access token.
Before you start
- Opencode installed, with
opencodeon your PATH. The setup script configures a harness; it does not install one. Install Opencode → - Node.js 18 or newer, which the setup script runs on.
- A Tokenmaxer API key. Create one →
Run the setup script
Run this, pick Opencode, and paste your key when it asks. It writes the configuration below for you, backing up anything it touches.
curl -fsSL https://tokenmaxer.dev/setup.sh | bashTo skip the menu, pass the harness: curl -fsSL https://tokenmaxer.dev/setup.sh | bash -s -- opencode. Run it again after connecting another provider plan to refresh the available models, or curl -fsSL https://tokenmaxer.dev/setup.sh | bash -s -- --undo to put your files back.
Test it
opencode run --model tokenmaxer/anthropic/claude-sonnet-5-5 'say hello'A reply, and `/models` lists the Tokenmaxer provider with every model you configured.
Configure it manually
This is what the setup script writes, for anyone who would rather edit the files themselves. Replace<YOUR_KEY> with your Tokenmaxer access token.
opencode takes the `/v1` suffix despite speaking the Anthropic API, because the AI SDK provider appends its own path segment.
Declare `limit` on every model
opencode has no catalogue entry for Tokenmaxer model ids and falls back to a default context window, which makes it compact far too early.
Declare `modalities` on every model
Same missing catalogue entry, and here the fallback is `false` for every modality. Drag an image or a PDF onto a model with no `modalities` and opencode replaces it with `ERROR: Cannot read … (this model does not support image input)` before the request leaves your machine, so it reads like the model refused rather than like a config gap. Every model except GPT-5.3 Codex Spark accepts images and PDFs.
Claude models
~/.config/opencode/opencode.jsonc
Adds Tokenmaxer as a provider. opencode deep-merges config, so your existing providers are untouched.
{
"provider": {
"tokenmaxer": {
"npm": "@ai-sdk/anthropic",
"name": "Tokenmaxer",
"options": {
"baseURL": "https://api.tokenmaxer.dev/v1",
"apiKey": "<YOUR_KEY>"
},
"models": {
"anthropic/claude-opus-5-5": {
"name": "Claude Opus 5.5",
"reasoning": true,
"tool_call": true,
"attachment": true,
"modalities": {
"input": [
"text",
"image",
"pdf"
],
"output": [
"text"
]
},
"limit": {
"context": 1000000,
"output": 128000
}
},
"anthropic/claude-sonnet-5-5": {
"name": "Claude Sonnet 5.5",
"reasoning": true,
"tool_call": true,
"attachment": true,
"modalities": {
"input": [
"text",
"image",
"pdf"
],
"output": [
"text"
]
},
"limit": {
"context": 1000000,
"output": 128000
}
},
"anthropic/claude-fable-5-1": {
"name": "Claude Fable 5.1",
"reasoning": true,
"tool_call": true,
"attachment": true,
"modalities": {
"input": [
"text",
"image",
"pdf"
],
"output": [
"text"
]
},
"limit": {
"context": 1000000,
"output": 64000
}
},
"anthropic/claude-haiku-5-5": {
"name": "Claude Haiku 5.5",
"reasoning": true,
"tool_call": true,
"attachment": true,
"modalities": {
"input": [
"text",
"image",
"pdf"
],
"output": [
"text"
]
},
"limit": {
"context": 1000000,
"output": 128000
}
}
}
}
}
}~/.config/opencode/opencode.jsonc
OptionalMakes Tokenmaxer the default. Skip it to keep your current default and pick Tokenmaxer from `/models` instead.
{
"model": "tokenmaxer/anthropic/claude-sonnet-5-5"
}ChatGPT models
~/.config/opencode/opencode.jsonc
Adds Tokenmaxer as a provider. opencode deep-merges config, so your existing providers are untouched.
{
"provider": {
"tokenmaxer-codex": {
"npm": "@ai-sdk/openai",
"name": "Tokenmaxer (Codex)",
"options": {
"baseURL": "https://api.tokenmaxer.dev/v1",
"apiKey": "<YOUR_KEY>"
},
"models": {
"openai/gpt-6.1-sol": {
"name": "GPT-6.1 Sol",
"reasoning": true,
"tool_call": true,
"attachment": true,
"modalities": {
"input": [
"text",
"image",
"pdf"
],
"output": [
"text"
]
},
"limit": {
"context": 872000,
"output": 128000
}
},
"openai/gpt-6-astra": {
"name": "GPT-6 Astra",
"reasoning": true,
"tool_call": true,
"attachment": true,
"modalities": {
"input": [
"text",
"image",
"pdf"
],
"output": [
"text"
]
},
"limit": {
"context": 872000,
"output": 128000
}
},
"openai/gpt-6-sol": {
"name": "GPT-6 Sol",
"reasoning": true,
"tool_call": true,
"attachment": true,
"modalities": {
"input": [
"text",
"image",
"pdf"
],
"output": [
"text"
]
},
"limit": {
"context": 872000,
"output": 128000
}
},
"openai/gpt-6-luna": {
"name": "GPT-6 Luna",
"reasoning": true,
"tool_call": true,
"attachment": true,
"modalities": {
"input": [
"text",
"image",
"pdf"
],
"output": [
"text"
]
},
"limit": {
"context": 872000,
"output": 128000
}
},
"openai/gpt-5.6-terra": {
"name": "GPT-5.6 Terra",
"reasoning": true,
"tool_call": true,
"attachment": true,
"modalities": {
"input": [
"text",
"image",
"pdf"
],
"output": [
"text"
]
},
"limit": {
"context": 872000,
"output": 128000
}
},
"openai/gpt-5.6-luna": {
"name": "GPT-5.6 Luna",
"reasoning": true,
"tool_call": true,
"attachment": true,
"modalities": {
"input": [
"text",
"image",
"pdf"
],
"output": [
"text"
]
},
"limit": {
"context": 872000,
"output": 128000
}
}
}
}
}
}~/.config/opencode/opencode.jsonc
OptionalMakes Tokenmaxer the default. Skip it to keep your current default and pick Tokenmaxer from `/models` instead.
{
"model": "tokenmaxer-codex/openai/gpt-5.6-terra"
}