Model Catalog
Every model CodeWork can run, what each one costs, and which thinking levels it supports.
CodeWork reads models from a catalog in your CodeWork home. The catalog lists each model’s provider, context window, input types, reasoning support, and price, and it’s what --provider, --model, and the model picker in your editor choose from.
codework models # every model, as provider/model
codework models --provider openai-codex # one provider
codework models providers # provider IDs onlyWhere the Catalog Comes From
| Provider | Source |
|---|---|
| OpenAI Codex | A fixed list maintained in CodeWork (below). |
| GitHub Copilot | models.dev, adjusted for Copilot’s endpoints and thinking levels. |
| Every other provider | models.dev: any provider whose models support tool calling. |
| Your own models | The models array in your settings, added to the catalog. |
The catalog is saved as ~/.codework/models.gen.json and refreshed when it’s older than 15 minutes. Change that window with CODEWORK_MODELS_TTL, refresh it by hand with codework models refresh --force, or pin it to a file you manage with CODEWORK_MODELS_FILE.
How Pricing Works
Prices are in US dollars per million tokens, with four rates per model:
| Rate | Charged for |
|---|---|
| Input | Prompt tokens the provider hasn’t cached |
| Output | Tokens the model generates, including reasoning tokens |
| Cache read | Prompt tokens served from the provider’s cache |
| Cache write | Prompt tokens written to the cache, on providers that bill for it |
Some models also have long-context tiers: once a request’s input passes a threshold, a higher set of rates applies to the whole request. They’re listed under cost.tiers in ~/.codework/models.gen.json.
Every codework run ends with a usage summary that adds these up for that run:
── usage ───────────────────────────────────────────────────────────────
model openai-codex/gpt-5.6-luna
tokens 6,172 input · 1,377 output · 412 reasoning · 19,581 total
cache 12,032 read · 0 write
cost $0.003127 · 3 turnsSubscriptions are billed by your plan
With OpenAI Codex or GitHub Copilot, the cost line is what the same tokens would cost at API rates. What you actually pay is set by your ChatGPT or Copilot plan. Use it to compare models, not as a bill.
OpenAI Codex Models
Available with codework auth login --openai-codex on a ChatGPT Plus or Pro plan. Prices are the API-equivalent rates CodeWork uses for the usage summary.
| Model | ID | Input | Cache Read | Cache Write | Output | Context |
|---|---|---|---|---|---|---|
| GPT-6 Luna | gpt-6-luna | $0.10 | $0.01 | $0.125 | $0.50 | 272K |
| GPT-5.6 Luna | gpt-5.6-luna | $0.20 | $0.02 | $0.25 | $1.20 | 272K |
| GPT-5.6 Terra | gpt-5.6-terra | $2.00 | $0.20 | $2.50 | $12.00 | 272K |
| GPT-5.6 Sol | gpt-5.6-sol | $5.00 | $0.50 | $6.25 | $30.00 | 272K |
| GPT-5.5 | gpt-5.5 | $5.00 | $0.50 | - | $30.00 | 272K |
| GPT-5.4 | gpt-5.4 | $2.50 | $0.25 | - | $15.00 | 272K |
| GPT-5.4 Mini | gpt-5.4-mini | $0.75 | $0.075 | - | $4.50 | 272K |
| GPT-5.3 Codex Spark | gpt-5.3-codex-spark | $1.75 | $0.175 | - | $14.00 | 128K |
Every Codex model reasons, accepts images (except Codex Spark, which is text only), and can write up to 128K output tokens. Requests served at the flex service tier cost half; priority costs 2× (2.5× for GPT-5.5).
Thinking Levels
Set the level with --thinking, thinkingLevel in settings, or the Thinking Effort picker in your editor. Ask for a level a model doesn’t support and CodeWork uses the next one up, or the highest it has.
| Model | Levels |
|---|---|
gpt-6-luna, gpt-5.6-luna | off, low, medium, high, xhigh, max |
gpt-5.6-sol, gpt-5.6-terra | low, medium, high, xhigh, max |
gpt-5.5 | off, low, medium, high, xhigh |
gpt-5.4, gpt-5.4-mini | low, medium, high, xhigh |
gpt-5.3-codex-spark | minimal, low, medium, high, xhigh |
Top OpenRouter Models
Available with OPENROUTER_API_KEY set. See OpenRouter. Prices come from models.dev and follow OpenRouter’s, so they can change between catalog refreshes.
| Model | ID | Input | Cache Read | Cache Write | Output | Context |
|---|---|---|---|---|---|---|
| Claude Opus 5.5 | anthropic/claude-opus-5.5 | $4.00 | $0.20 | $5.00 | $20.00 | 1M |
| Claude Sonnet 5.5 | anthropic/claude-sonnet-5.5 | $2.00 | $0.10 | $2.50 | $10.00 | 1M |
| GPT-6.1 Sol | openai/gpt-6.1-sol | $2.00 | $0.10 | $2.50 | $10.00 | 1.05M |
| Gemini 3.8 Flash | google/gemini-3.8-flash | $0.75 | $0.075 | $0.042 | $3.75 | 1M |
| Kimi K3 | moonshotai/kimi-k3 | $0.50 | $0.28 | - | $13.50 | 1M |
| DeepSeek V4.1 Flash | deepseek/deepseek-v4.1-flash | $0.30 | $0.006 | - | $1.20 | 1M |
| MiniMax M3 | minimax/minimax-m3 | $0.30 | $0.06 | - | $1.20 | 1M |
Use the ID with the openrouter provider, for example --provider openrouter --model deepseek/deepseek-v4.1-flash.