Token estimation
The token counter in the workbench tells you how much of your model's context window the current output will consume. It updates live as files come and go.
How the count is produced
Token counting runs against @dqbd/tiktoken, a WebAssembly build of
OpenAI's tokenizer. The default encoding is the one used by
o1-preview-2024-09-12, which most modern LLM models match closely
enough to be useful as a unified estimate. The result you see on screen
is the actual encoder output for the assembled blob, not a heuristic.
The encoder runs synchronously after the page boots, so toggling files on and off does not introduce a perceptible delay for typical projects.
The 1 MB cap
Inputs over 1 MB skip the encoder and fall back to
Math.ceil(length / 4). The reason lives in
apps/web/src/lib/tokens.ts: running the WASM encoder on multi-MB
blobs blocks the main thread long enough to feel like the tab has
frozen, and the approximation is good enough for a headline number on
inputs that size. The fallback also catches the case where the encoder
itself fails for any reason.
If you are working with a giant context and want an exact count, narrow the output with the filter patterns until you are under the cap, then read the precise number.
Model selector
The button next to the token counter opens a model selector. The
catalog ships with the build and is refreshed at every deploy from
models.dev, the same source the cost estimates use.
A few facts that make the picker behave the way it does:
- Models are sorted by release date descending, so the newest Claude, GPT, Gemini, and so on appear at the top.
- The picker shows the canonical model (e.g.
claude-sonnet-4-6) and avia {provider}hint for the cheapest priced provider that offers that model on the day of the deploy. If the cheapest provider changes between deploys, the hint changes; the canonical id stays the same so your selection survives. - The selected model's identifier is the catalog UID
(
lab/model-id), persisted in localStorage. Models removed from the catalog upstream are pruned at load time.
See also
- Token costs for which toggles cost or save tokens.
- Filter precedence for the layered filtering pipeline that decides which files contribute to the count.