diff --git a/app/lambdas/shared/config.py b/app/lambdas/shared/config.py
index 9796c2f..e240b5c 100644
--- a/app/lambdas/shared/config.py
+++ b/app/lambdas/shared/config.py
@@ -51,7 +51,7 @@ def _resolve_db_password() -> str:
 
 S3_BUCKET_NAME: str = os.environ.get("S3_BUCKET_NAME", "lauter-files")
 BEDROCK_MODEL_ID: str = os.environ.get(
-    "BEDROCK_MODEL_ID", "us.anthropic.claude-sonnet-4-6-v1:0"
+    "BEDROCK_MODEL_ID", "us.anthropic.claude-sonnet-4-6"
 )
 AWS_REGION: str = os.environ.get("AWS_REGION", "us-east-1")
 
diff --git a/app/lambdas/tests/test_recommendation_handler.py b/app/lambdas/tests/test_recommendation_handler.py
index daca11c..c45ef94 100644
--- a/app/lambdas/tests/test_recommendation_handler.py
+++ b/app/lambdas/tests/test_recommendation_handler.py
@@ -7,7 +7,7 @@
 os.environ.setdefault("DB_USERNAME", "test")
 os.environ.setdefault("DB_PASSWORD", "test")
 os.environ.setdefault("S3_BUCKET_NAME", "test-bucket")
-os.environ.setdefault("BEDROCK_MODEL_ID", "us.anthropic.claude-sonnet-4-6-v1:0")
+os.environ.setdefault("BEDROCK_MODEL_ID", "us.anthropic.claude-sonnet-4-6")
 os.environ.setdefault("AWS_REGION", "us-east-1")
 
 CV_ANALYSIS_RESULT = {
diff --git a/app/lambdas/tests/test_screening_eval_handler.py b/app/lambdas/tests/test_screening_eval_handler.py
index 80856d7..3b2072f 100644
--- a/app/lambdas/tests/test_screening_eval_handler.py
+++ b/app/lambdas/tests/test_screening_eval_handler.py
@@ -7,7 +7,7 @@
 os.environ.setdefault("DB_USERNAME", "test")
 os.environ.setdefault("DB_PASSWORD", "test")
 os.environ.setdefault("S3_BUCKET_NAME", "test-bucket")
-os.environ.setdefault("BEDROCK_MODEL_ID", "us.anthropic.claude-sonnet-4-6-v1:0")
+os.environ.setdefault("BEDROCK_MODEL_ID", "us.anthropic.claude-sonnet-4-6")
 os.environ.setdefault("AWS_REGION", "us-east-1")
 
 SAMPLE_RESULT = {
diff --git a/app/lambdas/tests/test_shared.py b/app/lambdas/tests/test_shared.py
index 8beea3b..7002669 100644
--- a/app/lambdas/tests/test_shared.py
+++ b/app/lambdas/tests/test_shared.py
@@ -9,7 +9,7 @@
 os.environ.setdefault("DB_USERNAME", "test")
 os.environ.setdefault("DB_PASSWORD", "test")
 os.environ.setdefault("S3_BUCKET_NAME", "test-bucket")
-os.environ.setdefault("BEDROCK_MODEL_ID", "us.anthropic.claude-sonnet-4-6-v1:0")
+os.environ.setdefault("BEDROCK_MODEL_ID", "us.anthropic.claude-sonnet-4-6")
 os.environ.setdefault("AWS_REGION", "us-east-1")
 
 
diff --git a/app/lambdas/tests/test_technical_eval_handler.py b/app/lambdas/tests/test_technical_eval_handler.py
index 2ed6fa6..e62516f 100644
--- a/app/lambdas/tests/test_technical_eval_handler.py
+++ b/app/lambdas/tests/test_technical_eval_handler.py
@@ -8,7 +8,7 @@
 os.environ.setdefault("DB_USERNAME", "test")
 os.environ.setdefault("DB_PASSWORD", "test")
 os.environ.setdefault("S3_BUCKET_NAME", "test-bucket")
-os.environ.setdefault("BEDROCK_MODEL_ID", "us.anthropic.claude-sonnet-4-6-v1:0")
+os.environ.setdefault("BEDROCK_MODEL_ID", "us.anthropic.claude-sonnet-4-6")
 os.environ.setdefault("AWS_REGION", "us-east-1")
 
 SAMPLE_RUBRIC_STRUCTURE = {
diff --git a/infra/ROLLOUT.md b/infra/ROLLOUT.md
index 90e87db..f516a5c 100644
--- a/infra/ROLLOUT.md
+++ b/infra/ROLLOUT.md
@@ -160,7 +160,7 @@ Edit `terraform.tfvars` with actual values:
 | `jwt_secret_key_arn` | `"arn:aws:secretsmanager:us-east-1:..."` | From Step 4 |
 | `enable_bedrock` | `false` | Set `true` after Step 2 when ready |
 | `alert_email` | `"team@provectus.com"` | CloudWatch alarm notifications (optional) |
-| `bedrock_model_id_heavy` | `"us.anthropic.claude-sonnet-4-6-v1:0"` | Heavy eval tasks (default works) |
+| `bedrock_model_id_heavy` | `"us.anthropic.claude-sonnet-4-6"` | Heavy eval tasks (default works) |
 | `bedrock_model_id_light` | `"us.anthropic.claude-haiku-4-5-20251001-v1:0"` | Light eval tasks (default works) |
 
 Sensitive variables can also be passed via env vars:
diff --git a/infra/terraform.tfvars.example b/infra/terraform.tfvars.example
index 5428240..2704e6b 100644
--- a/infra/terraform.tfvars.example
+++ b/infra/terraform.tfvars.example
@@ -44,6 +44,6 @@ alert_email = ""
 # Amazon Bedrock model IDs for evaluation Lambda functions
 # Ensure these models are enabled in your AWS account via the Bedrock console
 # Heavy: screening eval, technical eval, recommendation (complex reasoning)
-bedrock_model_id_heavy = "us.anthropic.claude-sonnet-4-6-v1:0"
+bedrock_model_id_heavy = "us.anthropic.claude-sonnet-4-6"
 # Light: CV analysis, feedback generation (simpler extraction/generation)
 bedrock_model_id_light = "us.anthropic.claude-haiku-4-5-20251001-v1:0"
diff --git a/infra/variables.tf b/infra/variables.tf
index 71eec5d..052bf63 100644
--- a/infra/variables.tf
+++ b/infra/variables.tf
@@ -87,7 +87,7 @@ variable "alert_email" {
 variable "bedrock_model_id_heavy" {
   description = "Bedrock model for complex evaluation tasks (screening, technical, recommendation)"
   type        = string
-  default     = "us.anthropic.claude-sonnet-4-6-v1:0"
+  default     = "us.anthropic.claude-sonnet-4-6"
 }
 
 variable "bedrock_model_id_light" {