@gtrabanco/pi-nan-provider 0.6.8 → 0.6.9

package/AGENTS.md

@@ -136,14 +136,21 @@ Every PR that changes code MUST bump `package.json` version in the same PR; CI p
   `supportsFinishReason: true` so pi-ai raises the retryable
   `Stream ended without finish_reason` (pi-ai's `RETRYABLE_PROVIDER_ERROR_PATTERN`
   matches `"ended without"`, so the turn is retried) instead of silently
-  synthesizing `stop`/`toolUse`. `supportsUsageInStreaming` stays `false` by
-  default (NaN's published schema does not document `stream_options`), but the
-  sanitizer gates its `stream_options` removal on the model's effective
-  `compat.supportsUsageInStreaming`, so a confirmed per-model `models.json`
-  override now yields real usage instead of being silently undone
-  (`test/issue-4-token-usage.test.ts`; issues #2, #4). Regression tests:
-  `test/issue-2-truncated-stream.test.ts`, `test/issue-4-token-usage.test.ts`;
-  issues #2 and #4.
+  synthesizing `stop`/`toolUse`. `supportsUsageInStreaming` is `true` for chat
+  models (issue #7): NaN's published schema does not document `stream_options`,
+  but the live gateway honors `stream_options.include_usage` — measured
+  2026-09-16 on `deepseek-v4-flash`, `glm5.3-flash`, `qwen3.6`, `mimo-v2.5` and
+  `gemma4` (0 usage chunks without the flag, exactly 1 with it, carrying
+  prompt/completion/reasoning/cached counts; a real pi session then recorded
+  real tokens where it recorded zeros). The sanitizer gates its
+  `stream_options` removal on the model's effective
+  `compat.supportsUsageInStreaming`, so the stock catalog reports real usage and
+  a per-model `models.json` override of `false` restores the strict payload
+  (`test/issue-4-token-usage.test.ts`, `test/issue-7-streaming-usage-default.test.ts`;
+  issues #2, #4, #7). Regression tests:
+  `test/issue-2-truncated-stream.test.ts`, `test/issue-4-token-usage.test.ts`,
+  `test/issue-7-streaming-usage-default.test.ts`;
+  issues #2, #4 and #7.
 - A NaN request that still exceeds the destination model's context window (the
   cross-model thinking guard is disabled with `NAN_THINKING_GUARD=0`, the
   inflation is not a `thinking` block, or the window is smaller) gets NaN's

package/README.es.md

@@ -71,14 +71,14 @@ Puedes sobrescribir el `compat` de cualquier modelo en `~/.pi/agent/models.json`
 
 > Poner `supportsFinishReason: false` restaura el antiguo stall silencioso — no recomendado.
 
-**Uso de tokens en streaming:** `supportsUsageInStreaming` es `false` por defecto porque el esquema publicado de NaN no documenta `stream_options`; sin él, el usage aparece a cero. Si has confirmado que tu modelo devuelve el chunk de usage en streaming, activalo por modelo — el sanitizer de peticiones entonces reenvía `stream_options: { "include_usage": true }` y pi muestra los tokens reales en lugar de ceros:
+**Uso de tokens en streaming:** `supportsUsageInStreaming` es `true` por defecto. El esquema publicado de NaN no documenta `stream_options`, pero el gateway real lo acepta y lo aplica — medido el 2026-09-16 ([#7](https://github.com/gtrabanco/pi-nan-provider/issues/7)): dos llamadas de streaming idénticas por modelo, 0 chunks de usage sin el flag y exactamente 1 con él, en `deepseek-v4-flash`, `glm5.3-flash`, `qwen3.6`, `mimo-v2.5` y `gemma4`. Por eso pi muestra tokens reales de entrada/salida/reasoning/caché en lugar de ceros. Si un modelo resulta no devolver el usage en streaming, desactívalo por modelo — el sanitizer entonces elimina `stream_options` y el payload vuelve a ser estricto:
 
 ```json
 {
   "providers": {
     "nan": {
       "modelOverrides": {
-        "qwen3.6": { "compat": { "supportsUsageInStreaming": true } }
+        "some-model": { "compat": { "supportsUsageInStreaming": false } }
       }
     }
   }

package/README.md

@@ -71,14 +71,14 @@ You can override any model's `compat` per-model in `~/.pi/agent/models.json` (pi
 
 > Setting `supportsFinishReason: false` restores the old silent-stall behavior — not recommended.
 
-**Streaming token usage:** `supportsUsageInStreaming` is `false` by default because NaN's published schema does not document `stream_options`; without it usage reads as zero. If you have confirmed that your model returns a streaming usage chunk, opt in per model — the request sanitizer then forwards `stream_options: { "include_usage": true }` and pi reports real token counts instead of zeros:
+**Streaming token usage:** `supportsUsageInStreaming` is `true` by default. NaN's published schema does not document `stream_options`, but the live gateway honors it — measured 2026-09-16 ([#7](https://github.com/gtrabanco/pi-nan-provider/issues/7)): two identical streaming calls per model, 0 usage chunks without the flag and exactly 1 with it, on `deepseek-v4-flash`, `glm5.3-flash`, `qwen3.6`, `mimo-v2.5` and `gemma4`. pi therefore reports real input/output/reasoning/cache token counts instead of zeros. If a model turns out not to report streaming usage, opt out per model — the request sanitizer then strips `stream_options` and the payload stays strict:
 
 ```json
 {
   "providers": {
     "nan": {
       "modelOverrides": {
-        "qwen3.6": { "compat": { "supportsUsageInStreaming": true } }
+        "some-model": { "compat": { "supportsUsageInStreaming": false } }
       }
     }
   }

package/package.json

@@ -1,6 +1,6 @@
 {
 	"name": "@gtrabanco/pi-nan-provider",
-	"version": "0.6.8",
+	"version": "0.6.9",
 	"description": "NaN Builders (api.nan.builders) model provider for pi - OpenAI-compatible registration with a models.dev-generated fallback, tier-aware live catalog, and MCP bridges (official web search + optional community media server)",
 	"keywords": [
 		"pi",

package/scripts/generate-models.ts

@@ -83,14 +83,16 @@ const LIVE_ONLY_MODEL_IDS: Record<string, string> = {
 const NAN_COMPAT = {
 	supportsDeveloperRole: false,
 	supportsReasoningEffort: true,
-	// pi-ai only sends `stream_options: { include_usage: true }` when this is
-	// not false. NaN's published schema does not document `stream_options`, so
-	// the conservative default is false: the sanitizer strips it and usage
-	// stays zero. A user who confirms their gateway reports streaming usage
-	// can opt in per model with a models.json compat override
-	// (`supportsUsageInStreaming: true`); the sanitizer then forwards
-	// `stream_options` instead of deleting it (issue #4, 2026-09-13).
-	supportsUsageInStreaming: false,
+	// NaN's published schema is silent about `stream_options`, but the live
+	// gateway honors it (issue #7, measured 2026-09-16 on five chat models:
+	// 0 usage chunks without the flag, exactly 1 with it). pi-ai only sends
+	// `stream_options: { include_usage: true }` when this is not false, and the
+	// sanitizer forwards it when the model declares true, so chat models opt in
+	// by default and pi reports real token counts instead of zeros (issue #4).
+	// A model that does not report streaming usage can still opt out per model
+	// with a models.json compat override (`supportsUsageInStreaming: false`);
+	// the sanitizer then strips `stream_options` and the payload stays strict.
+	supportsUsageInStreaming: true,
 	// The NaN/LiteLLM gateway intermittently closes SSE streams before emitting
 	// `finish_reason`. With true, pi-ai raises "Stream ended without
 	// finish_reason", which its retryable-provider pattern ("ended without")
@@ -102,7 +104,7 @@ const NAN_COMPAT = {
 };
 
 const NAN_COMPAT_NOTE =
-	"compat matches the maintainer's working ~/.pi/agent/models.json LiteLLM config for api.nan.builders (2026-09-04): supportsDeveloperRole false, supportsReasoningEffort true, maxTokensField max_tokens. NaN's docs example sets only supportsDeveloperRole: true and is not battle-tested. supportsFinishReason true (2026-09-13, issue #2): the LiteLLM gateway intermittently closes SSE streams before emitting finish_reason; with true pi-ai raises 'Stream ended without finish_reason', which matches pi-ai's retryable-provider pattern ('ended without') and is retried automatically, whereas false silently synthesized stop/toolUse and stalled the turn mid-answer. supportsUsageInStreaming false (2026-09-13): pi-ai only sends stream_options when this is not false, and NaN's published schema does not document it, so the conservative default stays false (sanitizer removes stream_options, usage reads zero). A user who has confirmed that their model returns a streaming usage chunk can opt in per model with a models.json compat override (supportsUsageInStreaming: true); the sanitizer then forwards stream_options instead of deleting it (issue #4).";
+	"compat matches the maintainer's working ~/.pi/agent/models.json LiteLLM config for api.nan.builders (2026-09-04): supportsDeveloperRole false, supportsReasoningEffort true, maxTokensField max_tokens. NaN's docs example sets only supportsDeveloperRole: true and is not battle-tested. supportsFinishReason true (2026-09-13, issue #2): the LiteLLM gateway intermittently closes SSE streams before emitting finish_reason; with true pi-ai raises 'Stream ended without finish_reason', which matches pi-ai's retryable-provider pattern ('ended without') and is retried automatically, whereas false silently synthesized stop/toolUse and stalled the turn mid-answer. supportsUsageInStreaming true (2026-09-16, issue #7): NaN's published schema is silent about stream_options, but the live gateway honors it — two identical streaming calls per model, differing only in stream_options: { include_usage: true }, returned 0 usage chunks without it and exactly 1 with it (prompt/completion/reasoning/cached token counts) on deepseek-v4-flash, glm5.3-flash, qwen3.6, mimo-v2.5 and gemma4, and a real pi session then recorded token counts where it recorded zeros. pi-ai only sends stream_options when this is not false, and the sanitizer forwards it when the model declares true, so chat models opt in by default and usage is reported (issue #4). A model that does not report streaming usage can still opt out per model with a models.json compat override (supportsUsageInStreaming: false); the sanitizer then strips stream_options and the payload stays strict.";
 
 interface ModelsDevModel {
 	id?: string;

package/scripts/models.generated.ts

@@ -1,7 +1,7 @@
 // This file is auto-generated by scripts/generate-models.ts
 // Do not edit manually — run `bun run generate-models` to update.
 //
-// Source: https://models.dev/api.json (provider "nan"), fetched 2026-09-14T20:40:55.922Z
+// Source: https://models.dev/api.json (provider "nan"), fetched 2026-09-16T19:40:56.569Z
 // Provenance: every contextWindow/maxTokens/input/cost value traces to
 // models.dev or to the per-entry notes below. Nothing is invented; entries
 // models.dev documents incompletely are omitted and flagged instead.
@@ -31,12 +31,12 @@ export const NAN_GENERATED_MODELS: readonly GeneratedModelEntry[] = [
 		"compat": {
 			"supportsDeveloperRole": false,
 			"supportsReasoningEffort": true,
-			"supportsUsageInStreaming": false,
+			"supportsUsageInStreaming": true,
 			"supportsFinishReason": true,
 			"maxTokensField": "max_tokens"
 		},
 		"notes": [
-			"compat matches the maintainer's working ~/.pi/agent/models.json LiteLLM config for api.nan.builders (2026-09-04): supportsDeveloperRole false, supportsReasoningEffort true, maxTokensField max_tokens. NaN's docs example sets only supportsDeveloperRole: true and is not battle-tested. supportsFinishReason true (2026-09-13, issue #2): the LiteLLM gateway intermittently closes SSE streams before emitting finish_reason; with true pi-ai raises 'Stream ended without finish_reason', which matches pi-ai's retryable-provider pattern ('ended without') and is retried automatically, whereas false silently synthesized stop/toolUse and stalled the turn mid-answer. supportsUsageInStreaming false (2026-09-13): pi-ai only sends stream_options when this is not false, and NaN's published schema does not document it, so the conservative default stays false (sanitizer removes stream_options, usage reads zero). A user who has confirmed that their model returns a streaming usage chunk can opt in per model with a models.json compat override (supportsUsageInStreaming: true); the sanitizer then forwards stream_options instead of deleting it (issue #4).",
+			"compat matches the maintainer's working ~/.pi/agent/models.json LiteLLM config for api.nan.builders (2026-09-04): supportsDeveloperRole false, supportsReasoningEffort true, maxTokensField max_tokens. NaN's docs example sets only supportsDeveloperRole: true and is not battle-tested. supportsFinishReason true (2026-09-13, issue #2): the LiteLLM gateway intermittently closes SSE streams before emitting finish_reason; with true pi-ai raises 'Stream ended without finish_reason', which matches pi-ai's retryable-provider pattern ('ended without') and is retried automatically, whereas false silently synthesized stop/toolUse and stalled the turn mid-answer. supportsUsageInStreaming true (2026-09-16, issue #7): NaN's published schema is silent about stream_options, but the live gateway honors it — two identical streaming calls per model, differing only in stream_options: { include_usage: true }, returned 0 usage chunks without it and exactly 1 with it (prompt/completion/reasoning/cached token counts) on deepseek-v4-flash, glm5.3-flash, qwen3.6, mimo-v2.5 and gemma4, and a real pi session then recorded token counts where it recorded zeros. pi-ai only sends stream_options when this is not false, and the sanitizer forwards it when the model declares true, so chat models opt in by default and usage is reported (issue #4). A model that does not report streaming usage can still opt out per model with a models.json compat override (supportsUsageInStreaming: false); the sanitizer then strips stream_options and the payload stays strict.",
 			"input includes image: NaN serves the Vision-Exp variant ('takes images as input', https://nan.builders/docs/models; the image_url content-parts in https://nan.builders/openapi.json list deepseek-v4-flash among the vision models). models.dev provider nan also lists text+image now (DeepSeek V4.1 Flash entry, checked 2026-09-13; its 2026-09-07 snapshot listed text only), so this override is kept as a pin for the vision capability rather than as a divergence."
 		],
 		"extras": {
@@ -92,12 +92,12 @@ export const NAN_GENERATED_MODELS: readonly GeneratedModelEntry[] = [
 		"compat": {
 			"supportsDeveloperRole": false,
 			"supportsReasoningEffort": true,
-			"supportsUsageInStreaming": false,
+			"supportsUsageInStreaming": true,
 			"supportsFinishReason": true,
 			"maxTokensField": "max_tokens"
 		},
 		"notes": [
-			"compat matches the maintainer's working ~/.pi/agent/models.json LiteLLM config for api.nan.builders (2026-09-04): supportsDeveloperRole false, supportsReasoningEffort true, maxTokensField max_tokens. NaN's docs example sets only supportsDeveloperRole: true and is not battle-tested. supportsFinishReason true (2026-09-13, issue #2): the LiteLLM gateway intermittently closes SSE streams before emitting finish_reason; with true pi-ai raises 'Stream ended without finish_reason', which matches pi-ai's retryable-provider pattern ('ended without') and is retried automatically, whereas false silently synthesized stop/toolUse and stalled the turn mid-answer. supportsUsageInStreaming false (2026-09-13): pi-ai only sends stream_options when this is not false, and NaN's published schema does not document it, so the conservative default stays false (sanitizer removes stream_options, usage reads zero). A user who has confirmed that their model returns a streaming usage chunk can opt in per model with a models.json compat override (supportsUsageInStreaming: true); the sanitizer then forwards stream_options instead of deleting it (issue #4)."
+			"compat matches the maintainer's working ~/.pi/agent/models.json LiteLLM config for api.nan.builders (2026-09-04): supportsDeveloperRole false, supportsReasoningEffort true, maxTokensField max_tokens. NaN's docs example sets only supportsDeveloperRole: true and is not battle-tested. supportsFinishReason true (2026-09-13, issue #2): the LiteLLM gateway intermittently closes SSE streams before emitting finish_reason; with true pi-ai raises 'Stream ended without finish_reason', which matches pi-ai's retryable-provider pattern ('ended without') and is retried automatically, whereas false silently synthesized stop/toolUse and stalled the turn mid-answer. supportsUsageInStreaming true (2026-09-16, issue #7): NaN's published schema is silent about stream_options, but the live gateway honors it — two identical streaming calls per model, differing only in stream_options: { include_usage: true }, returned 0 usage chunks without it and exactly 1 with it (prompt/completion/reasoning/cached token counts) on deepseek-v4-flash, glm5.3-flash, qwen3.6, mimo-v2.5 and gemma4, and a real pi session then recorded token counts where it recorded zeros. pi-ai only sends stream_options when this is not false, and the sanitizer forwards it when the model declares true, so chat models opt in by default and usage is reported (issue #4). A model that does not report streaming usage can still opt out per model with a models.json compat override (supportsUsageInStreaming: false); the sanitizer then strips stream_options and the payload stays strict."
 		],
 		"extras": {
 			"id": "gemma4",
@@ -155,18 +155,18 @@ export const NAN_GENERATED_MODELS: readonly GeneratedModelEntry[] = [
 		"compat": {
 			"supportsDeveloperRole": false,
 			"supportsReasoningEffort": true,
-			"supportsUsageInStreaming": false,
+			"supportsUsageInStreaming": true,
 			"supportsFinishReason": true,
 			"maxTokensField": "max_tokens"
 		},
 		"notes": [
-			"compat matches the maintainer's working ~/.pi/agent/models.json LiteLLM config for api.nan.builders (2026-09-04): supportsDeveloperRole false, supportsReasoningEffort true, maxTokensField max_tokens. NaN's docs example sets only supportsDeveloperRole: true and is not battle-tested. supportsFinishReason true (2026-09-13, issue #2): the LiteLLM gateway intermittently closes SSE streams before emitting finish_reason; with true pi-ai raises 'Stream ended without finish_reason', which matches pi-ai's retryable-provider pattern ('ended without') and is retried automatically, whereas false silently synthesized stop/toolUse and stalled the turn mid-answer. supportsUsageInStreaming false (2026-09-13): pi-ai only sends stream_options when this is not false, and NaN's published schema does not document it, so the conservative default stays false (sanitizer removes stream_options, usage reads zero). A user who has confirmed that their model returns a streaming usage chunk can opt in per model with a models.json compat override (supportsUsageInStreaming: true); the sanitizer then forwards stream_options instead of deleting it (issue #4)."
+			"compat matches the maintainer's working ~/.pi/agent/models.json LiteLLM config for api.nan.builders (2026-09-04): supportsDeveloperRole false, supportsReasoningEffort true, maxTokensField max_tokens. NaN's docs example sets only supportsDeveloperRole: true and is not battle-tested. supportsFinishReason true (2026-09-13, issue #2): the LiteLLM gateway intermittently closes SSE streams before emitting finish_reason; with true pi-ai raises 'Stream ended without finish_reason', which matches pi-ai's retryable-provider pattern ('ended without') and is retried automatically, whereas false silently synthesized stop/toolUse and stalled the turn mid-answer. supportsUsageInStreaming true (2026-09-16, issue #7): NaN's published schema is silent about stream_options, but the live gateway honors it — two identical streaming calls per model, differing only in stream_options: { include_usage: true }, returned 0 usage chunks without it and exactly 1 with it (prompt/completion/reasoning/cached token counts) on deepseek-v4-flash, glm5.3-flash, qwen3.6, mimo-v2.5 and gemma4, and a real pi session then recorded token counts where it recorded zeros. pi-ai only sends stream_options when this is not false, and the sanitizer forwards it when the model declares true, so chat models opt in by default and usage is reported (issue #4). A model that does not report streaming usage can still opt out per model with a models.json compat override (supportsUsageInStreaming: false); the sanitizer then strips stream_options and the payload stays strict."
 		],
 		"extras": {
 			"id": "glm5.3-flash",
 			"name": "GLM-5.3-Flash",
 			"description": "Native multimodal GLM model for efficient coding and long-horizon agent tasks",
-			"family": "glm",
+			"family": "glm-flash",
 			"attachment": true,
 			"reasoning": true,
 			"reasoning_options": [],
@@ -214,12 +214,12 @@ export const NAN_GENERATED_MODELS: readonly GeneratedModelEntry[] = [
 		"compat": {
 			"supportsDeveloperRole": false,
 			"supportsReasoningEffort": true,
-			"supportsUsageInStreaming": false,
+			"supportsUsageInStreaming": true,
 			"supportsFinishReason": true,
 			"maxTokensField": "max_tokens"
 		},
 		"notes": [
-			"compat matches the maintainer's working ~/.pi/agent/models.json LiteLLM config for api.nan.builders (2026-09-04): supportsDeveloperRole false, supportsReasoningEffort true, maxTokensField max_tokens. NaN's docs example sets only supportsDeveloperRole: true and is not battle-tested. supportsFinishReason true (2026-09-13, issue #2): the LiteLLM gateway intermittently closes SSE streams before emitting finish_reason; with true pi-ai raises 'Stream ended without finish_reason', which matches pi-ai's retryable-provider pattern ('ended without') and is retried automatically, whereas false silently synthesized stop/toolUse and stalled the turn mid-answer. supportsUsageInStreaming false (2026-09-13): pi-ai only sends stream_options when this is not false, and NaN's published schema does not document it, so the conservative default stays false (sanitizer removes stream_options, usage reads zero). A user who has confirmed that their model returns a streaming usage chunk can opt in per model with a models.json compat override (supportsUsageInStreaming: true); the sanitizer then forwards stream_options instead of deleting it (issue #4)."
+			"compat matches the maintainer's working ~/.pi/agent/models.json LiteLLM config for api.nan.builders (2026-09-04): supportsDeveloperRole false, supportsReasoningEffort true, maxTokensField max_tokens. NaN's docs example sets only supportsDeveloperRole: true and is not battle-tested. supportsFinishReason true (2026-09-13, issue #2): the LiteLLM gateway intermittently closes SSE streams before emitting finish_reason; with true pi-ai raises 'Stream ended without finish_reason', which matches pi-ai's retryable-provider pattern ('ended without') and is retried automatically, whereas false silently synthesized stop/toolUse and stalled the turn mid-answer. supportsUsageInStreaming true (2026-09-16, issue #7): NaN's published schema is silent about stream_options, but the live gateway honors it — two identical streaming calls per model, differing only in stream_options: { include_usage: true }, returned 0 usage chunks without it and exactly 1 with it (prompt/completion/reasoning/cached token counts) on deepseek-v4-flash, glm5.3-flash, qwen3.6, mimo-v2.5 and gemma4, and a real pi session then recorded token counts where it recorded zeros. pi-ai only sends stream_options when this is not false, and the sanitizer forwards it when the model declares true, so chat models opt in by default and usage is reported (issue #4). A model that does not report streaming usage can still opt out per model with a models.json compat override (supportsUsageInStreaming: false); the sanitizer then strips stream_options and the payload stays strict."
 		],
 		"extras": {
 			"id": "mimo-v2.5",
@@ -274,12 +274,12 @@ export const NAN_GENERATED_MODELS: readonly GeneratedModelEntry[] = [
 		"compat": {
 			"supportsDeveloperRole": false,
 			"supportsReasoningEffort": true,
-			"supportsUsageInStreaming": false,
+			"supportsUsageInStreaming": true,
 			"supportsFinishReason": true,
 			"maxTokensField": "max_tokens"
 		},
 		"notes": [
-			"compat matches the maintainer's working ~/.pi/agent/models.json LiteLLM config for api.nan.builders (2026-09-04): supportsDeveloperRole false, supportsReasoningEffort true, maxTokensField max_tokens. NaN's docs example sets only supportsDeveloperRole: true and is not battle-tested. supportsFinishReason true (2026-09-13, issue #2): the LiteLLM gateway intermittently closes SSE streams before emitting finish_reason; with true pi-ai raises 'Stream ended without finish_reason', which matches pi-ai's retryable-provider pattern ('ended without') and is retried automatically, whereas false silently synthesized stop/toolUse and stalled the turn mid-answer. supportsUsageInStreaming false (2026-09-13): pi-ai only sends stream_options when this is not false, and NaN's published schema does not document it, so the conservative default stays false (sanitizer removes stream_options, usage reads zero). A user who has confirmed that their model returns a streaming usage chunk can opt in per model with a models.json compat override (supportsUsageInStreaming: true); the sanitizer then forwards stream_options instead of deleting it (issue #4)."
+			"compat matches the maintainer's working ~/.pi/agent/models.json LiteLLM config for api.nan.builders (2026-09-04): supportsDeveloperRole false, supportsReasoningEffort true, maxTokensField max_tokens. NaN's docs example sets only supportsDeveloperRole: true and is not battle-tested. supportsFinishReason true (2026-09-13, issue #2): the LiteLLM gateway intermittently closes SSE streams before emitting finish_reason; with true pi-ai raises 'Stream ended without finish_reason', which matches pi-ai's retryable-provider pattern ('ended without') and is retried automatically, whereas false silently synthesized stop/toolUse and stalled the turn mid-answer. supportsUsageInStreaming true (2026-09-16, issue #7): NaN's published schema is silent about stream_options, but the live gateway honors it — two identical streaming calls per model, differing only in stream_options: { include_usage: true }, returned 0 usage chunks without it and exactly 1 with it (prompt/completion/reasoning/cached token counts) on deepseek-v4-flash, glm5.3-flash, qwen3.6, mimo-v2.5 and gemma4, and a real pi session then recorded token counts where it recorded zeros. pi-ai only sends stream_options when this is not false, and the sanitizer forwards it when the model declares true, so chat models opt in by default and usage is reported (issue #4). A model that does not report streaming usage can still opt out per model with a models.json compat override (supportsUsageInStreaming: false); the sanitizer then strips stream_options and the payload stays strict."
 		],
 		"extras": {
 			"id": "qwen3.6",
@@ -337,12 +337,12 @@ export const NAN_GENERATED_MODELS: readonly GeneratedModelEntry[] = [
 		"compat": {
 			"supportsDeveloperRole": false,
 			"supportsReasoningEffort": true,
-			"supportsUsageInStreaming": false,
+			"supportsUsageInStreaming": true,
 			"supportsFinishReason": true,
 			"maxTokensField": "max_tokens"
 		},
 		"notes": [
-			"compat matches the maintainer's working ~/.pi/agent/models.json LiteLLM config for api.nan.builders (2026-09-04): supportsDeveloperRole false, supportsReasoningEffort true, maxTokensField max_tokens. NaN's docs example sets only supportsDeveloperRole: true and is not battle-tested. supportsFinishReason true (2026-09-13, issue #2): the LiteLLM gateway intermittently closes SSE streams before emitting finish_reason; with true pi-ai raises 'Stream ended without finish_reason', which matches pi-ai's retryable-provider pattern ('ended without') and is retried automatically, whereas false silently synthesized stop/toolUse and stalled the turn mid-answer. supportsUsageInStreaming false (2026-09-13): pi-ai only sends stream_options when this is not false, and NaN's published schema does not document it, so the conservative default stays false (sanitizer removes stream_options, usage reads zero). A user who has confirmed that their model returns a streaming usage chunk can opt in per model with a models.json compat override (supportsUsageInStreaming: true); the sanitizer then forwards stream_options instead of deleting it (issue #4).",
+			"compat matches the maintainer's working ~/.pi/agent/models.json LiteLLM config for api.nan.builders (2026-09-04): supportsDeveloperRole false, supportsReasoningEffort true, maxTokensField max_tokens. NaN's docs example sets only supportsDeveloperRole: true and is not battle-tested. supportsFinishReason true (2026-09-13, issue #2): the LiteLLM gateway intermittently closes SSE streams before emitting finish_reason; with true pi-ai raises 'Stream ended without finish_reason', which matches pi-ai's retryable-provider pattern ('ended without') and is retried automatically, whereas false silently synthesized stop/toolUse and stalled the turn mid-answer. supportsUsageInStreaming true (2026-09-16, issue #7): NaN's published schema is silent about stream_options, but the live gateway honors it — two identical streaming calls per model, differing only in stream_options: { include_usage: true }, returned 0 usage chunks without it and exactly 1 with it (prompt/completion/reasoning/cached token counts) on deepseek-v4-flash, glm5.3-flash, qwen3.6, mimo-v2.5 and gemma4, and a real pi session then recorded token counts where it recorded zeros. pi-ai only sends stream_options when this is not false, and the sanitizer forwards it when the model declares true, so chat models opt in by default and usage is reported (issue #4). A model that does not report streaming usage can still opt out per model with a models.json compat override (supportsUsageInStreaming: false); the sanitizer then strips stream_options and the payload stays strict.",
 			"contextWindow 262,144: the earlier 1,000,000 override (maintainer-confirmed 2026-09-05) was withdrawn 2026-09-07 — the updated https://nan.builders/docs/models still states '262K token context, the model's native window' and models.dev agrees at 262,144; NaN docs are treated as the most reliable source (maintainer instruction, 2026-09-07)."
 		],
 		"extras": {
@@ -382,13 +382,13 @@ export const NAN_GENERATED_MODELS: readonly GeneratedModelEntry[] = [
 export const GENERATED_CATALOG_META = {
 	source: "https://models.dev/api.json",
 	modelsDevProvider: "nan",
-	fetchedAt: "2026-09-14T20:40:55.922Z",
+	fetchedAt: "2026-09-16T19:40:56.569Z",
 	modelCount: 6,
 	models: ["deepseek-v4-flash","gemma4","glm5.3-flash","mimo-v2.5","qwen3.6","qwen3.8-flash"],
 	notes: [
 	"live-only: \"glm5.3\" kept out of the static catalog (premium-tier model (models.dev now documents it with 1M context / 131,072 max output; NaN docs https://nan.builders/docs/models + https://nan.builders/openapi.json, checked 2026-09-13) kept live-only so a non-premium key never sees a model it cannot call when the live /models fetch is unavailable; premium keys still get it via the /models refresh with conservative placeholder limits)",
 	"provider-removed: \"glm5.2\" excluded from the catalog (removed by NaN (2026-09-05); absent from the official chat model list in https://nan.builders/openapi.json and https://nan.builders/docs/models (checked 2026-09-07) while models.dev provider nan still listed it — excluded so regeneration does not resurrect it)",
-	"compat matches the maintainer's working ~/.pi/agent/models.json LiteLLM config for api.nan.builders (2026-09-04): supportsDeveloperRole false, supportsReasoningEffort true, maxTokensField max_tokens. NaN's docs example sets only supportsDeveloperRole: true and is not battle-tested. supportsFinishReason true (2026-09-13, issue #2): the LiteLLM gateway intermittently closes SSE streams before emitting finish_reason; with true pi-ai raises 'Stream ended without finish_reason', which matches pi-ai's retryable-provider pattern ('ended without') and is retried automatically, whereas false silently synthesized stop/toolUse and stalled the turn mid-answer. supportsUsageInStreaming false (2026-09-13): pi-ai only sends stream_options when this is not false, and NaN's published schema does not document it, so the conservative default stays false (sanitizer removes stream_options, usage reads zero). A user who has confirmed that their model returns a streaming usage chunk can opt in per model with a models.json compat override (supportsUsageInStreaming: true); the sanitizer then forwards stream_options instead of deleting it (issue #4).",
+	"compat matches the maintainer's working ~/.pi/agent/models.json LiteLLM config for api.nan.builders (2026-09-04): supportsDeveloperRole false, supportsReasoningEffort true, maxTokensField max_tokens. NaN's docs example sets only supportsDeveloperRole: true and is not battle-tested. supportsFinishReason true (2026-09-13, issue #2): the LiteLLM gateway intermittently closes SSE streams before emitting finish_reason; with true pi-ai raises 'Stream ended without finish_reason', which matches pi-ai's retryable-provider pattern ('ended without') and is retried automatically, whereas false silently synthesized stop/toolUse and stalled the turn mid-answer. supportsUsageInStreaming true (2026-09-16, issue #7): NaN's published schema is silent about stream_options, but the live gateway honors it — two identical streaming calls per model, differing only in stream_options: { include_usage: true }, returned 0 usage chunks without it and exactly 1 with it (prompt/completion/reasoning/cached token counts) on deepseek-v4-flash, glm5.3-flash, qwen3.6, mimo-v2.5 and gemma4, and a real pi session then recorded token counts where it recorded zeros. pi-ai only sends stream_options when this is not false, and the sanitizer forwards it when the model declares true, so chat models opt in by default and usage is reported (issue #4). A model that does not report streaming usage can still opt out per model with a models.json compat override (supportsUsageInStreaming: false); the sanitizer then strips stream_options and the payload stays strict.",
 	"input includes image: NaN serves the Vision-Exp variant ('takes images as input', https://nan.builders/docs/models; the image_url content-parts in https://nan.builders/openapi.json list deepseek-v4-flash among the vision models). models.dev provider nan also lists text+image now (DeepSeek V4.1 Flash entry, checked 2026-09-13; its 2026-09-07 snapshot listed text only), so this override is kept as a pin for the vision capability rather than as a divergence.",
 	"contextWindow 262,144: the earlier 1,000,000 override (maintainer-confirmed 2026-09-05) was withdrawn 2026-09-07 — the updated https://nan.builders/docs/models still states '262K token context, the model's native window' and models.dev agrees at 262,144; NaN docs are treated as the most reliable source (maintainer instruction, 2026-09-07)."
 ],

package/src/fetch-models.ts

@@ -177,10 +177,11 @@ export function mergeLiveWithGenerated(
 			// gateway cuts SSE streams before finish_reason, so supportsFinishReason
 			// must stay true (pi-ai then raises the retryable "Stream ended without
 			// finish_reason" instead of silently stalling), and
-			// supportsUsageInStreaming stays false because NaN's schema does not
-			// document `stream_options`: the sanitizer strips it unless the model's
-			// effective compat opts in (issue #4), so false matches the default wire
-			// payload.
+			// supportsUsageInStreaming is true because the live gateway honors
+			// `stream_options.include_usage` even though the published schema is
+			// silent (issue #7, measured 2026-09-16): the sanitizer forwards the
+			// field and pi reports real token counts. A model can still opt out
+			// with a models.json override of false.
 			models.push({
 				id,
 				name: id,
@@ -192,7 +193,7 @@ export function mergeLiveWithGenerated(
 				cost: { input: 0, output: 0, cacheRead: 0, cacheWrite: 0 },
 				contextWindow: UNKNOWN_MODEL_LIMITS.contextWindow,
 				maxTokens: UNKNOWN_MODEL_LIMITS.maxTokens,
-				compat: { supportsFinishReason: true, supportsUsageInStreaming: false },
+				compat: { supportsFinishReason: true, supportsUsageInStreaming: true },
 			});
 			unknown.push(id);
 		}

package/src/openai-compat-sanitizer.ts

@@ -33,14 +33,16 @@
  *     not silently lost.
  *  4. Top-level fields NaN's schema does not list: `store` and
  *     `stream_options`. These are opt-in/usage fields pi-ai sends by default
- *     for a "standard" provider; NaN does not document them, so they are
- *     removed by default. `stream_options` is the one exception that can be
- *     explicitly opted into: when the model's effective
- *     `compat.supportsUsageInStreaming` is true (a catalog flag or a user
- *     `models.json` override), the caller has confirmed the gateway reports
- *     usage and `stream_options` is preserved — deleting it unconditionally
- *     would silently zero out `message.usage` (issue #4). `store` is always
- *     removed.
+ *     for a "standard" provider; NaN's published schema does not document
+ *     them, so they are removed by default. `stream_options` is the one
+ *     exception that can be explicitly opted into: when the model's effective
+ *     `compat.supportsUsageInStreaming` is true (the catalog default for chat
+ *     models since issue #7 — the live gateway was measured honoring
+ *     `include_usage` on 2026-09-16 — or a user `models.json` override), the
+ *     gateway reports usage and `stream_options` is preserved — deleting it
+ *     unconditionally would silently zero out `message.usage` (issue #4).
+ *     A per-model override of `false` keeps the strict payload. `store` is
+ *     always removed.
  *  5. An EMPTY `tools` array. Verified against the live gateway (2026-09-09):
  *     NaN rejects `tools: []` with the same 400, while `stream: true`, a
  *     `system` message, string content, and a `tool` role message are all

package/src/provider-factory.ts

@@ -88,16 +88,17 @@ export async function resolveOpenAICompletionsApi(): Promise<OpenAICompletionsAp
  * ./openai-compat-sanitizer.ts). NaN returns HTTP 400 `Invalid request. Check
  * your request parameters.` for any payload that violates it — including a
  * replayed assistant message with a `toolCall` block inside `content`, a
- * `reasoning_details` field, or undocumented top-level fields like `store` /
- * `stream_options`. Sanitizing via the `onPayload` hook works regardless of
+ * `reasoning_details` field, or the undocumented top-level `store` field.
+ * Sanitizing via the `onPayload` hook works regardless of
  * which pi-ai version the runtime bundles, so the fix is not tied to a
  * specific upstream build.
  *
  * `stream_options` is the one field whose removal is conditional: when the
- * model's effective `compat.supportsUsageInStreaming` is true (catalog value
- * or user `models.json` override), pi-ai requested usage and the gateway will
- * return it — stripping the field would silently zero `message.usage`
- * (issue #4). Every other model keeps the strict payload.
+ * model's effective `compat.supportsUsageInStreaming` is true (the catalog
+ * default for chat models since issue #7, or a user `models.json` override),
+ * pi-ai requested usage and the gateway will return it — stripping the field
+ * would silently zero `message.usage` (issue #4). Every other model keeps the
+ * strict payload.
  *
  * Any caller-supplied `onPayload` (e.g. pi's own debug/passthrough hook) is
  * preserved and chained AFTER sanitization, so the final payload is always