From 8a568aeacc53c8b1f704ecd6877d3560bab4a5c8 Mon Sep 17 00:00:00 2001
From: Michael Neale <michael.neale@gmail.com>
Date: Sun, 31 May 2026 10:30:04 +1000
Subject: [PATCH] fix(canonical): restore GPT-4o and Azure GPT-3.5-turbo output
 limits

Address codex review feedback on #9484 (P2).

The bulk canonical regeneration in this branch flipped the bare
`openai/gpt-4o` alias from the 2024-08-06 snapshot to the older
2024-05-13 snapshot, dropping limit.output from 16384 to 4096 (and
reverting pricing). Since ModelConfig::with_canonical_limits copies
limit.output into max_tokens when GOOSE_MAX_TOKENS is unset, ordinary
gpt-4o responses would be capped at 4096 and truncate output/tool calls.

Restore openai/gpt-4o to the current 2024-08-06 values (output 16384,
input 2.5 / output 10.0 / cache_read 1.25). Apply the same fix to
azure/gpt-3.5-turbo, which the regeneration regressed the same way
(0613 -> 0301 snapshot, context/output 16384 -> 4096).

These were the only native-provider (openai/anthropic/google/azure)
output-limit downgrades introduced by the regeneration; remaining
changes are upstream drift on third-party router prefixes.

Signed-off-by: Michael Neale <michael.neale@gmail.com>
---
 .../canonical/data/canonical_models.json      | 27 ++++++++++---------
 1 file changed, 14 insertions(+), 13 deletions(-)

diff --git a/crates/goose/src/providers/canonical/data/canonical_models.json b/crates/goose/src/providers/canonical/data/canonical_models.json
index 7bd11e6772..9bef411a86 100644
--- a/crates/goose/src/providers/canonical/data/canonical_models.json
+++ b/crates/goose/src/providers/canonical/data/canonical_models.json
@@ -18107,15 +18107,15 @@
   },
   {
     "id": "azure/gpt-3.5-turbo",
-    "name": "GPT-3.5 Turbo 0301",
+    "name": "GPT-3.5 Turbo 0613",
     "family": "gpt",
     "attachment": false,
     "reasoning": false,
     "tool_call": false,
     "temperature": true,
     "knowledge": "2021-08",
-    "release_date": "2023-03-01",
-    "last_updated": "2023-03-01",
+    "release_date": "2023-06-13",
+    "last_updated": "2023-06-13",
     "modalities": {
       "input": [
         "text"
@@ -18126,12 +18126,12 @@
     },
     "open_weights": false,
     "cost": {
-      "input": 1.5,
-      "output": 2.0
+      "input": 3.0,
+      "output": 4.0
     },
     "limit": {
-      "context": 4096,
-      "output": 4096
+      "context": 16384,
+      "output": 16384
     }
   },
   {
@@ -91587,15 +91587,15 @@
   },
   {
     "id": "openai/gpt-4o",
-    "name": "GPT-4o (2024-05-13)",
+    "name": "GPT-4o (2024-08-06)",
     "family": "gpt",
     "attachment": true,
     "reasoning": false,
     "tool_call": true,
     "temperature": true,
     "knowledge": "2023-09",
-    "release_date": "2024-05-13",
-    "last_updated": "2024-05-13",
+    "release_date": "2024-08-06",
+    "last_updated": "2024-08-06",
     "modalities": {
       "input": [
         "text",
@@ -91607,12 +91607,13 @@
     },
     "open_weights": false,
     "cost": {
-      "input": 5.0,
-      "output": 15.0
+      "input": 2.5,
+      "output": 10.0,
+      "cache_read": 1.25
     },
     "limit": {
       "context": 128000,
-      "output": 4096
+      "output": 16384
     }
   },
   {