Skip to content

fix(models): refresh model metadata so newer models expose thinking levels #97

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
SamYue1 wants to merge 1 commit into from
Closed

fix(models): refresh model metadata so newer models expose thinking levels #97

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
2 changes: 2 additions & 0 deletions CHANGELOG.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,6 +2,8 @@

## Unreleased

- Refresh the model capability metadata so model ids added since the last refresh report their image, reasoning, and effort support instead of falling back to the text-only, non-reasoning defaults. `deepseek/deepseek-v4.1-flash`, `gpt-6-astra`, and `inclusionai/ling-3.0-flash-sante:free` now expose the thinking levels the public model reference documents, and no longer drop `reasoning_effort` from requests.
- Drop the manual effort overrides for the Meta Muse Spark models: the refreshed catalog carries those levels itself, and the table is documented to lose entries once that happens.
## 0.6.4 - 2026-09-03

- Refresh the generated Command Code capability catalog from `command-code@1.40.1` to `command-code@1.44.0`, adding current image-input, reasoning, effort, and output-limit metadata for newly published models.
Expand Down
2 changes: 1 addition & 1 deletion README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -139,7 +139,7 @@ The following environment variables are intended for tests, local mocks, and com

## Image input

The provider advertises image input only for models marked with the `image` input modality in the official Command Code CLI model catalog. The capability snapshot currently follows `command-code@1.44.0`; unknown models default to text-only until their upstream metadata is reviewed. A daily GitHub Actions job synchronizes the CLI version, image capabilities, reasoning flags, reasoning efforts, and model-specific output limits with the latest published CLI package and opens or updates a reviewable pull request when they change. Pricing remains manually reviewed because temporary promotions and long-context tiers require explicit review.
The provider advertises image input only for models marked with the `image` input modality in the official Command Code CLI model catalog. The capability snapshot currently follows `command-code@1.54.0`; unknown models default to text-only until their upstream metadata is reviewed. A daily GitHub Actions job synchronizes the CLI version, image capabilities, reasoning flags, reasoning efforts, and model-specific output limits with the latest published CLI package and opens or updates a reviewable pull request when they change. Pricing remains manually reviewed because temporary promotions and long-context tiers require explicit review.

For vision-capable models, Pi's native provider adapters forward image blocks from user messages and tool results using the documented OpenAI or Anthropic message schema. Unknown and text-only models remain marked text-only in Pi.

Expand Down
10 changes: 3 additions & 7 deletions src/commandcode-catalog-overrides.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,11 +15,7 @@ import type { CommandCodeReasoningEffort } from "./commandcode-catalog.ts"
export const MODEL_EFFORT_OVERRIDES: Readonly<
Record<string, readonly CommandCodeReasoningEffort[]>
> = {
// Meta Muse Spark: the CLI ships no effort levels, but the endpoint accepts
// `reasoning_effort` for these models and other hosts expose the same set.
"meta/muse-spark-1.1": ["minimal", "low", "medium", "high", "xhigh"],
"meta/muse-spark-1.2": ["minimal", "low", "medium", "high", "xhigh"],
"meta/muse-spark-1.2-contributor": ["minimal", "low", "medium", "high", "xhigh"],
"meta/muse-spark-1.3": ["minimal", "low", "medium", "high", "xhigh"],
"meta/muse-spark-1.3-contributor": ["minimal", "low", "medium", "high", "xhigh"],
// Empty since command-code@1.54.0: the CLI catalog publishes effort levels itself
// for every model it lists, including the Meta Muse Spark models this table used
// to carry (muse-spark-1.1/1.2/1.2-contributor/1.3/1.3-contributor).
}
19 changes: 17 additions & 2 deletions src/commandcode-catalog.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,10 +1,10 @@
export const COMMAND_CODE_CLI_VERSION = "1.44.0"
export const COMMAND_CODE_CLI_VERSION = "1.54.0"

export type CommandCodeInputType = "text" | "image"
export type CommandCodeReasoningEffort = "minimal" | "low" | "medium" | "high" | "xhigh" | "max"

/**
* Generated from command-code@1.44.0 by `npm run sync:commandcode-catalog`.
* Generated from command-code@1.54.0 by `npm run sync:commandcode-catalog`.
* Do not edit manually.
*/
export const MODEL_INPUT_MODALITIES: Readonly<Record<string, readonly CommandCodeInputType[]>> = {
Expand All @@ -17,6 +17,7 @@ export const MODEL_INPUT_MODALITIES: Readonly<Record<string, readonly CommandCod
"claude-sonnet-4-6": ["text", "image"],
"claude-sonnet-5": ["text", "image"],
"deepseek/deepseek-v4-flash-vision-exp": ["text", "image"],
"deepseek/deepseek-v4.1-flash": ["text", "image"],
"google/gemini-3.1-flash-lite": ["text", "image"],
"google/gemini-3.5-flash": ["text", "image"],
"google/gemini-3.5-flash-lite": ["text", "image"],
Expand All @@ -30,6 +31,7 @@ export const MODEL_INPUT_MODALITIES: Readonly<Record<string, readonly CommandCod
"gpt-5.6-luna": ["text", "image"],
"gpt-5.6-sol": ["text", "image"],
"gpt-5.6-terra": ["text", "image"],
"gpt-6-astra": ["text", "image"],
"meta/muse-spark-1.1": ["text", "image"],
"meta/muse-spark-1.2": ["text", "image"],
"meta/muse-spark-1.2-contributor": ["text", "image"],
Expand All @@ -53,6 +55,7 @@ export const MODEL_INPUT_MODALITIES: Readonly<Record<string, readonly CommandCod
"thinkingmachines/inkling": ["text", "image"],
"thinkingmachines/inkling-small": ["text", "image"],
"xai/grok-4.5": ["text", "image"],
"xai/grok-4.6": ["text", "image"],
"xiaomi/mimo-v2.5": ["text", "image"],
"z-ai/glm-5.3-flash": ["text", "image"],
}
Expand All @@ -69,6 +72,7 @@ export const MODEL_REASONING: Readonly<Record<string, true>> = {
"deepseek/deepseek-v4-flash-fast": true,
"deepseek/deepseek-v4-flash-vision-exp": true,
"deepseek/deepseek-v4-pro": true,
"deepseek/deepseek-v4.1-flash": true,
"google/gemini-3.1-flash-lite": true,
"google/gemini-3.5-flash": true,
"google/gemini-3.5-flash-lite": true,
Expand All @@ -82,6 +86,8 @@ export const MODEL_REASONING: Readonly<Record<string, true>> = {
"gpt-5.6-luna": true,
"gpt-5.6-sol": true,
"gpt-5.6-terra": true,
"gpt-6-astra": true,
"inclusionai/ling-3.0-flash-sante:free": true,
"meituan/LongCat-2.0:free": true,
"meta/muse-spark-1.1": true,
"meta/muse-spark-1.2": true,
Expand Down Expand Up @@ -129,6 +135,7 @@ export const MODEL_EFFORTS: Readonly<Record<string, readonly CommandCodeReasonin
"deepseek/deepseek-v4-flash-fast": ["low", "high", "max"],
"deepseek/deepseek-v4-flash-vision-exp": ["high", "max"],
"deepseek/deepseek-v4-pro": ["high", "max"],
"deepseek/deepseek-v4.1-flash": ["low", "high", "max"],
"google/gemini-3.1-flash-lite": ["low", "medium", "high"],
"google/gemini-3.5-flash": ["low", "medium", "high"],
"google/gemini-3.5-flash-lite": ["low", "medium", "high"],
Expand All @@ -142,6 +149,13 @@ export const MODEL_EFFORTS: Readonly<Record<string, readonly CommandCodeReasonin
"gpt-5.6-luna": ["low", "medium", "high", "xhigh", "max"],
"gpt-5.6-sol": ["low", "medium", "high", "xhigh", "max"],
"gpt-5.6-terra": ["low", "medium", "high", "xhigh", "max"],
"gpt-6-astra": ["low", "medium", "high", "xhigh", "max"],
"meta/muse-spark-1.1": ["low", "medium", "high", "xhigh"],
"meta/muse-spark-1.2": ["low", "medium", "high", "xhigh"],
"meta/muse-spark-1.2-contributor": ["low", "medium", "high", "xhigh"],
"meta/muse-spark-1.3": ["low", "medium", "high", "xhigh", "max"],
"meta/muse-spark-1.3-contributor": ["low", "medium", "high", "xhigh"],
"MiniMaxAI/MiniMax-M3": ["low", "medium", "high"],
"moonshotai/Kimi-K3": ["low", "high", "max"],
"Qwen/Qwen3.8-27B": ["low", "medium", "xhigh"],
"Qwen/Qwen3.8-Flash": ["low", "medium", "xhigh"],
Expand All @@ -157,6 +171,7 @@ export const MODEL_EFFORTS: Readonly<Record<string, readonly CommandCodeReasonin
}

export const MODEL_MAX_OUTPUT_TOKENS: Readonly<Record<string, number>> = {
"inclusionai/ling-3.0-flash-sante:free": 32_768,
"poolside/laguna-s-2.1-free": 32_768,
"Qwen/Qwen3.8-27B": 32_768,
"z-ai/glm-5.3-flash": 131_072,
Expand Down
1 change: 0 additions & 1 deletion tests/test-models.ts
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -198,7 +198,6 @@ describe("commandCodeModelsFromApiResponse()", () => {

it("merges manual effort overrides over the generated catalog", () => {
const validEfforts = new Set(["minimal", "low", "medium", "high", "xhigh", "max"])
assert.ok(Object.keys(MODEL_EFFORT_OVERRIDES).length > 0)
for (const [modelId, efforts] of Object.entries(MODEL_EFFORT_OVERRIDES)) {
assert.equal(MODEL_REASONING[modelId], true, `${modelId} override needs a reasoning flag`)
assert.equal(
Expand Down