diff --git a/docs/my-website/docs/proxy/team_model_overrides.md b/docs/my-website/docs/proxy/team_model_overrides.md
new file mode 100644
index 00000000000..0af49400c14
--- /dev/null
+++ b/docs/my-website/docs/proxy/team_model_overrides.md
@@ -0,0 +1,297 @@
+import Tabs from '@theme/Tabs';
+import TabItem from '@theme/TabItem';
+
+# Team Model Overrides (Per-Member Model Access)
+
+Grant different team members access to different models within the same team.
+
+By default, all team members share the same `team.models` list. With **team model overrides**, you can:
+
+- Set `default_models` on a team — a base set of models every member gets
+- Set per-member `models` overrides — additional models for specific members
+- A member's **effective models** = `default_models` + their personal `models` override
+
+:::info
+
+This feature is gated behind a feature flag. Set this in your environment or proxy config:
+
+```
+LITELLM_TEAM_MODEL_OVERRIDES=true
+```
+
+Teams that don't use `default_models` or member overrides are completely unaffected.
+
+:::
+
+## Quick Start
+
+### 1. Enable the feature flag
+
+<Tabs>
+<TabItem value="env" label="Environment Variable">
+
+```bash
+export LITELLM_TEAM_MODEL_OVERRIDES=true
+```
+
+</TabItem>
+
+<TabItem value="yaml" label="Proxy Config (YAML)">
+
+```yaml
+environment_variables:
+  LITELLM_TEAM_MODEL_OVERRIDES: "true"
+```
+
+</TabItem>
+</Tabs>
+
+### 2. Create a team with `default_models`
+
+`default_models` is the base set of models all team members get access to. It must be a subset of `team.models`.
+
+```shell
+curl -X POST 'http://0.0.0.0:4000/team/new' \
+  -H 'Authorization: Bearer <your-master-key>' \
+  -H 'Content-Type: application/json' \
+  -d '{
+    "team_alias": "ml-team",
+    "models": ["gpt-4o", "gpt-4o-mini", "claude-sonnet"],
+    "default_models": ["gpt-4o"]
+  }'
+```
+
+| Field | Description |
+| --- | --- |
+| `models` | Full list of models the team is allowed to use (same as existing behavior) |
+| `default_models` | Subset of `models` that every member gets by default |
+
+:::info
+
+`default_models` must be a subset of `team.models`. Setting `default_models: ["claude-opus"]` when `models: ["gpt-4o"]` will return a 400 error.
+
+:::
+
+### 3. Add a member to the team
+
+```shell
+curl -X POST 'http://0.0.0.0:4000/team/member_add' \
+  -H 'Authorization: Bearer <your-master-key>' \
+  -H 'Content-Type: application/json' \
+  -d '{
+    "team_id": "<team-id>",
+    "member": {"role": "user", "user_id": "user-1"}
+  }'
+```
+
+At this point, `user-1` can access the `default_models` (`gpt-4o`).
+
+### 4. Set per-member model overrides
+
+Grant `user-1` access to `claude-sonnet` in addition to the team defaults:
+
+```shell
+curl -X POST 'http://0.0.0.0:4000/team/member_update' \
+  -H 'Authorization: Bearer <your-master-key>' \
+  -H 'Content-Type: application/json' \
+  -d '{
+    "team_id": "<team-id>",
+    "user_id": "user-1",
+    "models": ["claude-sonnet"]
+  }'
+```
+
+Now `user-1`'s **effective models** = `default_models` + member `models` = `["gpt-4o", "claude-sonnet"]`.
+
+:::info
+
+Member `models` must be a subset of `team.models`. You cannot grant a member access to models the team itself doesn't have.
+
+:::
+
+### 5. Generate a key and make requests
+
+```shell
+curl -X POST 'http://0.0.0.0:4000/key/generate' \
+  -H 'Authorization: Bearer <your-master-key>' \
+  -H 'Content-Type: application/json' \
+  -d '{
+    "team_id": "<team-id>",
+    "user_id": "user-1"
+  }'
+```
+
+The key will automatically inherit the user's effective models.
+
+<Tabs>
+<TabItem label="Allowed (gpt-4o)" value="allowed-default">
+
+`gpt-4o` is in `default_models` — allowed for all team members.
+
+```shell
+curl -X POST 'http://0.0.0.0:4000/chat/completions' \
+  -H 'Authorization: Bearer <user-key>' \
+  -H 'Content-Type: application/json' \
+  -d '{
+    "model": "gpt-4o",
+    "messages": [{"role": "user", "content": "Hello"}]
+  }'
+```
+
+</TabItem>
+
+<TabItem label="Allowed (claude-sonnet)" value="allowed-member">
+
+`claude-sonnet` is in this member's `models` override — allowed for this specific user.
+
+```shell
+curl -X POST 'http://0.0.0.0:4000/chat/completions' \
+  -H 'Authorization: Bearer <user-key>' \
+  -H 'Content-Type: application/json' \
+  -d '{
+    "model": "claude-sonnet",
+    "messages": [{"role": "user", "content": "Hello"}]
+  }'
+```
+
+</TabItem>
+
+<TabItem label="Blocked (gpt-4o-mini)" value="blocked">
+
+`gpt-4o-mini` is in `team.models` but NOT in this member's effective models — blocked.
+
+```shell
+curl -X POST 'http://0.0.0.0:4000/chat/completions' \
+  -H 'Authorization: Bearer <user-key>' \
+  -H 'Content-Type: application/json' \
+  -d '{
+    "model": "gpt-4o-mini",
+    "messages": [{"role": "user", "content": "Hello"}]
+  }'
+```
+
+Expected error:
+```json
+{
+  "error": {
+    "message": "model not allowed for this key",
+    "type": "key_model_access_denied",
+    "code": "401"
+  }
+}
+```
+
+</TabItem>
+</Tabs>
+
+## How It Works
+
+```
+team.models         = [gpt-4o, gpt-4o-mini, claude-sonnet]   # full team allowlist
+team.default_models = [gpt-4o]                                 # base for all members
+
+member.models       = [claude-sonnet]                          # per-user override
+
+effective_models    = default_models ∪ member.models
+                    = [gpt-4o, claude-sonnet]                   # what this member can use
+```
+
+| Concept | Description |
+| --- | --- |
+| `team.models` | The full set of models the team can access. Acts as the ceiling — no member can exceed this. |
+| `team.default_models` | Base models every member gets. Must be a subset of `team.models`. |
+| `member.models` | Additional models for a specific member. Must be a subset of `team.models`. |
+| **Effective models** | `default_models ∪ member.models` — the actual models a member can use at runtime. |
+
+## Key Generation Behavior
+
+When generating a key for a team member:
+
+| Scenario | Behavior |
+| --- | --- |
+| Key with no `models` specified | Key inherits the member's effective models |
+| Key with `models` that are a subset of effective models | Key is created with only the requested models |
+| Key with `models` outside effective models | Returns 403 — model not available for this user |
+
+```shell
+# This works — gpt-4o is in effective models
+curl -X POST 'http://0.0.0.0:4000/key/generate' \
+  -H 'Authorization: Bearer <your-master-key>' \
+  -H 'Content-Type: application/json' \
+  -d '{
+    "team_id": "<team-id>",
+    "user_id": "user-1",
+    "models": ["gpt-4o"]
+  }'
+
+# This fails 403 — gpt-4o-mini is NOT in effective models
+curl -X POST 'http://0.0.0.0:4000/key/generate' \
+  -H 'Authorization: Bearer <your-master-key>' \
+  -H 'Content-Type: application/json' \
+  -d '{
+    "team_id": "<team-id>",
+    "user_id": "user-1",
+    "models": ["gpt-4o-mini"]
+  }'
+```
+
+## Budget Preservation
+
+Updating a member's `models` does not affect their budget or rate limits. You can safely update models without worrying about losing existing budget configuration:
+
+```shell
+# Set budget
+curl -X POST 'http://0.0.0.0:4000/team/member_update' \
+  -H 'Authorization: Bearer <your-master-key>' \
+  -H 'Content-Type: application/json' \
+  -d '{
+    "team_id": "<team-id>",
+    "user_id": "user-1",
+    "max_budget_in_team": 100.0,
+    "tpm_limit": 5000
+  }'
+
+# Update models later — budget is preserved
+curl -X POST 'http://0.0.0.0:4000/team/member_update' \
+  -H 'Authorization: Bearer <your-master-key>' \
+  -H 'Content-Type: application/json' \
+  -d '{
+    "team_id": "<team-id>",
+    "user_id": "user-1",
+    "models": ["claude-sonnet", "gpt-4o"]
+  }'
+```
+
+## Teams Without Overrides
+
+Teams that don't set `default_models` or member `models` are **completely unaffected**. The existing `team.models` behavior continues to work as before — all members share the same model list.
+
+```shell
+# Normal team — no overrides, existing behavior preserved
+curl -X POST 'http://0.0.0.0:4000/team/new' \
+  -H 'Authorization: Bearer <your-master-key>' \
+  -H 'Content-Type: application/json' \
+  -d '{
+    "team_alias": "simple-team",
+    "models": ["gpt-4o", "gpt-4o-mini"]
+  }'
+```
+
+All members of this team can access `gpt-4o` and `gpt-4o-mini` — no change from existing behavior.
+
+## Database Migration
+
+The feature adds two new columns:
+
+| Table | Column | Type | Default |
+| --- | --- | --- | --- |
+| `LiteLLM_TeamTable` | `default_models` | `String[]` | `[]` |
+| `LiteLLM_TeamMembership` | `models` | `String[]` | `[]` |
+
+- **Default deployments**: Columns are auto-created on startup via `prisma db push`.
+- **`disable_prisma_schema_update: true` deployments**: Apply the migration manually before enabling the feature flag:
+
+```sql
+ALTER TABLE "LiteLLM_TeamTable" ADD COLUMN IF NOT EXISTS "default_models" TEXT[] DEFAULT ARRAY[]::TEXT[];
+ALTER TABLE "LiteLLM_TeamMembership" ADD COLUMN IF NOT EXISTS "models" TEXT[] DEFAULT ARRAY[]::TEXT[];
+```
diff --git a/docs/my-website/sidebars.js b/docs/my-website/sidebars.js
index c8ebb110c59..287206bc456 100644
--- a/docs/my-website/sidebars.js
+++ b/docs/my-website/sidebars.js
@@ -486,7 +486,8 @@ const sidebars = {
             "proxy/model_access",
             "proxy/model_access_groups",
             "proxy/access_groups",
-            "proxy/team_model_add"
+            "proxy/team_model_add",
+            "proxy/team_model_overrides"
           ]
         },
         {