diff --git a/providers/sap-ai-core/models/amazon--nova-lite.toml b/providers/sap-ai-core/models/amazon--nova-lite.toml
new file mode 100644
index 0000000000..c5e32ccc10
--- /dev/null
+++ b/providers/sap-ai-core/models/amazon--nova-lite.toml
@@ -0,0 +1,21 @@
+name = "amazon--nova-lite"
+description = "Amazon Nova Lite model"
+release_date = "2026-07-07"
+last_updated = "2026-07-07"
+attachment = true
+reasoning = false
+temperature = true
+tool_call = true
+open_weights = false
+
+[cost]
+input = 0.04
+output = 0.16
+
+[limit]
+context = 1000000
+output = 4096
+
+[modalities]
+input = ["text", "image"]
+output = ["text"]
diff --git a/providers/sap-ai-core/models/amazon--nova-micro.toml b/providers/sap-ai-core/models/amazon--nova-micro.toml
new file mode 100644
index 0000000000..7703ded224
--- /dev/null
+++ b/providers/sap-ai-core/models/amazon--nova-micro.toml
@@ -0,0 +1,21 @@
+name = "amazon--nova-micro"
+description = "Amazon Nova Micro model"
+release_date = "2026-07-07"
+last_updated = "2026-07-07"
+attachment = false
+reasoning = false
+temperature = true
+tool_call = true
+open_weights = false
+
+[cost]
+input = 0.03
+output = 0.1
+
+[limit]
+context = 128000
+output = 4096
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/sap-ai-core/models/amazon--nova-pro.toml b/providers/sap-ai-core/models/amazon--nova-pro.toml
new file mode 100644
index 0000000000..a7358e8c0e
--- /dev/null
+++ b/providers/sap-ai-core/models/amazon--nova-pro.toml
@@ -0,0 +1,21 @@
+name = "amazon--nova-pro"
+description = "Amazon Nova Pro model"
+release_date = "2026-07-07"
+last_updated = "2026-07-07"
+attachment = true
+reasoning = false
+temperature = true
+tool_call = true
+open_weights = false
+
+[cost]
+input = 0.56
+output = 2.13
+
+[limit]
+context = 300000
+output = 4096
+
+[modalities]
+input = ["text", "image"]
+output = ["text"]
diff --git a/providers/sap-ai-core/models/amazon--titan-embed-text.toml b/providers/sap-ai-core/models/amazon--titan-embed-text.toml
new file mode 100644
index 0000000000..2a664c0675
--- /dev/null
+++ b/providers/sap-ai-core/models/amazon--titan-embed-text.toml
@@ -0,0 +1,21 @@
+name = "amazon--titan-embed-text"
+description = "Amazon titan-embed-text model"
+release_date = "2026-07-07"
+last_updated = "2026-07-07"
+attachment = false
+reasoning = false
+temperature = false
+tool_call = false
+open_weights = false
+
+[cost]
+input = 0.14
+output = 0
+
+[limit]
+context = 131072
+output = 4096
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/sap-ai-core/models/anthropic--claude-3-haiku.toml b/providers/sap-ai-core/models/anthropic--claude-3-haiku.toml
index 311362a2a4..56d90b59f1 100644
--- a/providers/sap-ai-core/models/anthropic--claude-3-haiku.toml
+++ b/providers/sap-ai-core/models/anthropic--claude-3-haiku.toml
@@ -1,24 +1,19 @@
 name = "anthropic--claude-3-haiku"
-description = "Fast Claude model for responsive assistance, classification, and lightweight agents"
-family = "claude-haiku"
-release_date = "2024-03-13"
+base_model = "anthropic/claude-3-haiku-20240307"
 last_updated = "2024-03-13"
 attachment = true
-reasoning = false
-temperature = true
-knowledge = "2023-08-31"
 tool_call = true
 open_weights = false
 
 [cost]
-input = 0.25
-output = 1.25
+input = 0.24
+output = 0.89
 cache_read = 0.03
-cache_write = 0.30
+cache_write = 0.3
 
 [limit]
-context = 200_000
-output = 4_096
+context = 200000
+output = 4096
 
 [modalities]
 input = ["text", "image", "pdf"]
diff --git a/providers/sap-ai-core/models/anthropic--claude-3-opus.toml b/providers/sap-ai-core/models/anthropic--claude-3-opus.toml
index 6a0213fb03..62a59f51f8 100644
--- a/providers/sap-ai-core/models/anthropic--claude-3-opus.toml
+++ b/providers/sap-ai-core/models/anthropic--claude-3-opus.toml
@@ -3,22 +3,23 @@ description = "Flagship Claude model for deep reasoning, coding, and long-horizo
 family = "claude-opus"
 release_date = "2024-02-29"
 last_updated = "2024-02-29"
+status = "deprecated"
 attachment = true
 reasoning = false
 temperature = true
-knowledge = "2023-08-31"
 tool_call = true
+knowledge = "2023-08-31"
 open_weights = false
 
 [cost]
-input = 15.00
-output = 75.00
-cache_read = 1.50
+input = 15
+output = 75
+cache_read = 1.5
 cache_write = 18.75
 
 [limit]
-context = 200_000
-output = 4_096
+context = 200000
+output = 4096
 
 [modalities]
 input = ["text", "image", "pdf"]
diff --git a/providers/sap-ai-core/models/anthropic--claude-3-sonnet.toml b/providers/sap-ai-core/models/anthropic--claude-3-sonnet.toml
index 5adb9bfa72..0c6d3297c5 100644
--- a/providers/sap-ai-core/models/anthropic--claude-3-sonnet.toml
+++ b/providers/sap-ai-core/models/anthropic--claude-3-sonnet.toml
@@ -3,22 +3,23 @@ description = "Balanced Claude model for coding, analysis, agent workflows, and
 family = "claude-sonnet"
 release_date = "2024-03-04"
 last_updated = "2024-03-04"
+status = "deprecated"
 attachment = true
 reasoning = false
 temperature = true
-knowledge = "2023-08-31"
 tool_call = true
+knowledge = "2023-08-31"
 open_weights = false
 
 [cost]
-input = 3.00
-output = 15.00
-cache_read = 0.30
+input = 3
+output = 15
+cache_read = 0.3
 cache_write = 3.75
 
 [limit]
-context = 200_000
-output = 4_096
+context = 200000
+output = 4096
 
 [modalities]
 input = ["text", "image", "pdf"]
diff --git a/providers/sap-ai-core/models/anthropic--claude-3.5-sonnet.toml b/providers/sap-ai-core/models/anthropic--claude-3.5-sonnet.toml
index c78a092e76..47d14d825d 100644
--- a/providers/sap-ai-core/models/anthropic--claude-3.5-sonnet.toml
+++ b/providers/sap-ai-core/models/anthropic--claude-3.5-sonnet.toml
@@ -1,25 +1,15 @@
 name = "anthropic--claude-3.5-sonnet"
-description = "Balanced Claude model for coding, analysis, agent workflows, and cost control"
-family = "claude-sonnet"
-release_date = "2024-10-22"
+base_model = "anthropic/claude-3-5-sonnet-20241022"
 last_updated = "2024-10-22"
-attachment = true
-reasoning = false
-temperature = true
-tool_call = true
-knowledge = "2024-04-30"
+status = "deprecated"
 open_weights = false
 
 [cost]
-input = 3.00
-output = 15.00
-cache_read = 0.30
+input = 3
+output = 15
+cache_read = 0.3
 cache_write = 3.75
 
-[limit]
-context = 200_000
-output = 8_192
-
 [modalities]
 input = ["text", "image", "pdf"]
 output = ["text"]
diff --git a/providers/sap-ai-core/models/anthropic--claude-3.7-sonnet.toml b/providers/sap-ai-core/models/anthropic--claude-3.7-sonnet.toml
index 66b71a3adc..7b8d7f3e9f 100644
--- a/providers/sap-ai-core/models/anthropic--claude-3.7-sonnet.toml
+++ b/providers/sap-ai-core/models/anthropic--claude-3.7-sonnet.toml
@@ -1,27 +1,17 @@
 name = "anthropic--claude-3.7-sonnet"
-description = "Balanced Claude model for coding, analysis, agent workflows, and cost control"
-family = "claude-sonnet"
-release_date = "2025-02-24"
+base_model = "anthropic/claude-3-7-sonnet-20250219"
 last_updated = "2025-02-24"
-attachment = true
-reasoning = true
+status = "deprecated"
 # Bedrock Invoke body: $.thinking = {"type":"enabled","budget_tokens":N} or {"type":"disabled"}, with N >= 1024 and N < $.max_tokens. Bedrock Converse nests that object at $.additionalModelRequestFields.thinking. https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference.html#converse-additional-model-request-fields (accessed 2026-06-25)
-reasoning_options = [{ type = "budget_tokens", min = 1_024 }]
-temperature = true
-tool_call = true
-knowledge = "2024-10-31"
+reasoning_options = [{ type = "budget_tokens", min = 1024 }]
 open_weights = false
 
 [cost]
-input = 3.00
-output = 15.00
-cache_read = 0.30
+input = 3
+output = 15
+cache_read = 0.3
 cache_write = 3.75
 
-[limit]
-context = 200_000
-output = 64_000
-
 [modalities]
 input = ["text", "image", "pdf"]
 output = ["text"]
diff --git a/providers/sap-ai-core/models/anthropic--claude-4-opus.toml b/providers/sap-ai-core/models/anthropic--claude-4-opus.toml
index 99505a6b2c..b0947cabba 100644
--- a/providers/sap-ai-core/models/anthropic--claude-4-opus.toml
+++ b/providers/sap-ai-core/models/anthropic--claude-4-opus.toml
@@ -3,23 +3,25 @@ description = "Flagship Claude model for deep reasoning, coding, and long-horizo
 family = "claude-opus"
 release_date = "2025-05-22"
 last_updated = "2025-05-22"
+status = "deprecated"
 attachment = true
 reasoning = true
-reasoning_options = [{ type = "budget_tokens", min = 1_024 }]
+# Bedrock Invoke body: $.thinking = {"type":"enabled","budget_tokens":N} or {"type":"disabled"}, with N >= 1024 and N < $.max_tokens. Bedrock Converse nests that object at $.additionalModelRequestFields.thinking. https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference.html#converse-additional-model-request-fields (accessed 2026-06-25)
+reasoning_options = [{ type = "budget_tokens", min = 1024 }]
 temperature = true
 tool_call = true
 knowledge = "2025-03-31"
 open_weights = false
 
 [cost]
-input = 15.00
-output = 75.00
-cache_read = 1.50
+input = 15
+output = 75
+cache_read = 1.5
 cache_write = 18.75
 
 [limit]
-context = 200_000
-output = 32_000
+context = 200000
+output = 32000
 
 [modalities]
 input = ["text", "image", "pdf"]
diff --git a/providers/sap-ai-core/models/anthropic--claude-4-sonnet.toml b/providers/sap-ai-core/models/anthropic--claude-4-sonnet.toml
index c4a7eaaebd..ce037fdd0e 100644
--- a/providers/sap-ai-core/models/anthropic--claude-4-sonnet.toml
+++ b/providers/sap-ai-core/models/anthropic--claude-4-sonnet.toml
@@ -1,25 +1,26 @@
 name = "anthropic--claude-4-sonnet"
-description = "Balanced Claude model for coding, analysis, agent workflows, and cost control"
+description = "Anthropic Claude 4 Sonnet for coding and agent capabilities"
 family = "claude-sonnet"
 release_date = "2025-05-22"
 last_updated = "2025-05-22"
 attachment = true
 reasoning = true
-reasoning_options = [{ type = "budget_tokens", min = 1_024 }]
+# Bedrock Invoke body: $.thinking = {"type":"enabled","budget_tokens":N} or {"type":"disabled"}, with N >= 1024 and N < $.max_tokens. Bedrock Converse nests that object at $.additionalModelRequestFields.thinking. https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference.html#converse-additional-model-request-fields (accessed 2026-06-25)
+reasoning_options = [{ type = "budget_tokens", min = 1024 }]
 temperature = true
 tool_call = true
 knowledge = "2025-03-31"
 open_weights = false
 
 [cost]
-input = 3.00
-output = 15.00
-cache_read = 0.30
-cache_write = 3.75
+input = 2.04
+output = 9.88
+cache_read = 0.2
+cache_write = 2.54
 
 [limit]
-context = 200_000
-output = 64_000
+context = 200000
+output = 64000
 
 [modalities]
 input = ["text", "image", "pdf"]
diff --git a/providers/sap-ai-core/models/anthropic--claude-4.5-haiku.toml b/providers/sap-ai-core/models/anthropic--claude-4.5-haiku.toml
index 757858ba75..f0db91500f 100644
--- a/providers/sap-ai-core/models/anthropic--claude-4.5-haiku.toml
+++ b/providers/sap-ai-core/models/anthropic--claude-4.5-haiku.toml
@@ -1,25 +1,26 @@
 name = "anthropic--claude-4.5-haiku"
-description = "Fast Claude model for responsive assistance, classification, and lightweight agents"
+description = "Anthropic Claude 4.5 Haiku"
 family = "claude-haiku"
 release_date = "2025-10-15"
 last_updated = "2025-10-15"
 attachment = true
 reasoning = true
-reasoning_options = [{ type = "budget_tokens", min = 1_024 }]
+# Bedrock Invoke body: $.thinking = {"type":"enabled","budget_tokens":N} or {"type":"disabled"}, with N >= 1024 and N < $.max_tokens. Bedrock Converse nests that object at $.additionalModelRequestFields.thinking. https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference.html#converse-additional-model-request-fields (accessed 2026-06-25)
+reasoning_options = [{ type = "budget_tokens", min = 1024 }]
 temperature = true
-knowledge = "2025-02-28"
 tool_call = true
+knowledge = "2025-02-28"
 open_weights = false
 
 [cost]
-input = 1.00
-output = 5.00
-cache_read = 0.10
-cache_write = 1.25
+input = 0.79
+output = 3.67
+cache_read = 0.08
+cache_write = 0.99
 
 [limit]
-context = 200_000
-output = 64_000
+context = 200000
+output = 64000
 
 [modalities]
 input = ["text", "image", "pdf"]
diff --git a/providers/sap-ai-core/models/anthropic--claude-4.5-opus.toml b/providers/sap-ai-core/models/anthropic--claude-4.5-opus.toml
index 0b527c74f1..9081e555c7 100644
--- a/providers/sap-ai-core/models/anthropic--claude-4.5-opus.toml
+++ b/providers/sap-ai-core/models/anthropic--claude-4.5-opus.toml
@@ -1,25 +1,26 @@
 name = "anthropic--claude-4.5-opus"
-description = "Flagship Claude model for deep reasoning, coding, and long-horizon agents"
+description = "Anthropic Claude 4.5 Opus"
 family = "claude-opus"
 release_date = "2025-11-24"
 last_updated = "2025-11-24"
 attachment = true
 reasoning = true
-reasoning_options = [{ type = "budget_tokens", min = 1_024 }]
+# Bedrock Invoke body: $.thinking = {"type":"enabled","budget_tokens":N} or {"type":"disabled"}, with N >= 1024 and N < $.max_tokens. Bedrock Converse nests that object at $.additionalModelRequestFields.thinking. https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference.html#converse-additional-model-request-fields (accessed 2026-06-25)
+reasoning_options = [{ type = "budget_tokens", min = 1024 }]
 temperature = true
 tool_call = true
 knowledge = "2025-05"
 open_weights = false
 
 [cost]
-input = 5.00
-output = 25.00
-cache_read = 0.50
-cache_write = 6.25
+input = 3.67
+output = 18.06
+cache_read = 0.37
+cache_write = 4.59
 
 [limit]
-context = 200_000
-output = 64_000
+context = 200000
+output = 64000
 
 [modalities]
 input = ["text", "image", "pdf"]
diff --git a/providers/sap-ai-core/models/anthropic--claude-4.5-sonnet.toml b/providers/sap-ai-core/models/anthropic--claude-4.5-sonnet.toml
index 29940237c9..1acf216ceb 100644
--- a/providers/sap-ai-core/models/anthropic--claude-4.5-sonnet.toml
+++ b/providers/sap-ai-core/models/anthropic--claude-4.5-sonnet.toml
@@ -1,25 +1,26 @@
 name = "anthropic--claude-4.5-sonnet"
-description = "Balanced Claude model for coding, analysis, agent workflows, and cost control"
+description = "Anthropic Claude 4.5 Sonnet"
 family = "claude-sonnet"
 release_date = "2025-09-29"
 last_updated = "2025-09-29"
 attachment = true
 reasoning = true
-reasoning_options = [{ type = "budget_tokens", min = 1_024 }]
+# Bedrock Invoke body: $.thinking = {"type":"enabled","budget_tokens":N} or {"type":"disabled"}, with N >= 1024 and N < $.max_tokens. Bedrock Converse nests that object at $.additionalModelRequestFields.thinking. https://docs.aws.amazon.com/bedrock/latest/userguide/conversation-inference.html#converse-additional-model-request-fields (accessed 2026-06-25)
+reasoning_options = [{ type = "budget_tokens", min = 1024 }]
 temperature = true
 tool_call = true
 knowledge = "2025-07-31"
 open_weights = false
 
 [cost]
-input = 3.00
-output = 15.00
-cache_read = 0.30
-cache_write = 3.75
+input = 2.23
+output = 10.87
+cache_read = 0.2
+cache_write = 2.54
 
 [limit]
-context = 200_000
-output = 64_000
+context = 200000
+output = 64000
 
 [modalities]
 input = ["text", "image", "pdf"]
diff --git a/providers/sap-ai-core/models/anthropic--claude-4.6-opus.toml b/providers/sap-ai-core/models/anthropic--claude-4.6-opus.toml
index 8c0ccd4d6a..ce276c108a 100644
--- a/providers/sap-ai-core/models/anthropic--claude-4.6-opus.toml
+++ b/providers/sap-ai-core/models/anthropic--claude-4.6-opus.toml
@@ -1,26 +1,26 @@
 name = "anthropic--claude-4.6-opus"
-description = "Flagship Claude model for deep reasoning, coding, and long-horizon agents"
+description = "Anthropic Claude 4.6 Opus"
 family = "claude-opus"
 release_date = "2026-02-05"
 last_updated = "2026-03-13"
 attachment = true
 reasoning = true
 # Bedrock Claude 4.6 prefers $.thinking.type = "adaptive" plus $.output_config.effort = "low"|"medium"|"high"|"max"; manual enabled budget_tokens >= 1024 is deprecated. Converse prefixes both paths with $.additionalModelRequestFields. https://docs.anthropic.com/en/docs/build-with-claude/extended-thinking (accessed 2026-06-25)
-reasoning_options = [{ type = "effort", values = ["low", "medium", "high", "max"] }, { type = "budget_tokens", min = 1_024 }]
+reasoning_options = [{ type = "effort", values = ["low", "medium", "high", "max"] }, { type = "budget_tokens", min = 1024 }]
 temperature = true
 tool_call = true
 knowledge = "2025-05"
 open_weights = false
 
 [cost]
-input = 5.00
-output = 25.00
-cache_read = 0.50
-cache_write = 6.25
+input = 3.67
+output = 18.06
+cache_read = 0.37
+cache_write = 4.59
 
 [limit]
-context = 1_000_000
-output = 128_000
+context = 200000
+output = 128000
 
 [modalities]
 input = ["text", "image", "pdf"]
diff --git a/providers/sap-ai-core/models/anthropic--claude-4.6-sonnet.toml b/providers/sap-ai-core/models/anthropic--claude-4.6-sonnet.toml
index a08489ddb6..b60514f3a9 100644
--- a/providers/sap-ai-core/models/anthropic--claude-4.6-sonnet.toml
+++ b/providers/sap-ai-core/models/anthropic--claude-4.6-sonnet.toml
@@ -1,26 +1,26 @@
 name = "anthropic--claude-4.6-sonnet"
-description = "Balanced Claude model for coding, analysis, agent workflows, and cost control"
+description = "Anthropic Claude 4.6 Sonnet"
 family = "claude-sonnet"
 release_date = "2026-02-17"
 last_updated = "2026-03-13"
 attachment = true
 reasoning = true
 # Bedrock Claude 4.6 prefers $.thinking.type = "adaptive" plus $.output_config.effort = "low"|"medium"|"high"; manual enabled budget_tokens >= 1024 is deprecated. Converse prefixes both paths with $.additionalModelRequestFields. https://docs.anthropic.com/en/docs/build-with-claude/extended-thinking (accessed 2026-06-25)
-reasoning_options = [{ type = "effort", values = ["low", "medium", "high"] }, { type = "budget_tokens", min = 1_024 }]
+reasoning_options = [{ type = "effort", values = ["low", "medium", "high"] }, { type = "budget_tokens", min = 1024 }]
 temperature = true
 tool_call = true
 knowledge = "2025-08"
 open_weights = false
 
 [cost]
-input = 3.00
-output = 15.00
-cache_read = 0.30
-cache_write = 3.75
+input = 2.23
+output = 10.87
+cache_read = 0.2
+cache_write = 2.54
 
 [limit]
-context = 1_000_000
-output = 64_000
+context = 1000000
+output = 64000
 
 [modalities]
 input = ["text", "image", "pdf"]
diff --git a/providers/sap-ai-core/models/anthropic--claude-4.7-opus.toml b/providers/sap-ai-core/models/anthropic--claude-4.7-opus.toml
index c11a5e84c5..b37a0f26ff 100644
--- a/providers/sap-ai-core/models/anthropic--claude-4.7-opus.toml
+++ b/providers/sap-ai-core/models/anthropic--claude-4.7-opus.toml
@@ -1,5 +1,5 @@
 name = "anthropic--claude-4.7-opus"
-description = "Flagship Claude model for deep reasoning, coding, and long-horizon agents"
+description = "Anthropic Claude 4.7 Opus"
 family = "claude-opus"
 release_date = "2026-04-16"
 last_updated = "2026-04-16"
@@ -13,14 +13,14 @@ knowledge = "2026-01-31"
 open_weights = false
 
 [cost]
-input = 5.00
-output = 25.00
-cache_read = 0.50
-cache_write = 6.25
+input = 3.67
+output = 18.06
+cache_read = 0.37
+cache_write = 4.59
 
 [limit]
-context = 1_000_000
-output = 128_000
+context = 1000000
+output = 128000
 
 [modalities]
 input = ["text", "image", "pdf"]
diff --git a/providers/sap-ai-core/models/cohere--command-a-reasoning.toml b/providers/sap-ai-core/models/cohere--command-a-reasoning.toml
new file mode 100644
index 0000000000..361ed43314
--- /dev/null
+++ b/providers/sap-ai-core/models/cohere--command-a-reasoning.toml
@@ -0,0 +1,21 @@
+name = "cohere--command-a-reasoning"
+description = "Cohere's Command A Reasoning model"
+release_date = "2026-07-07"
+last_updated = "2026-07-07"
+attachment = false
+reasoning = false
+temperature = true
+tool_call = true
+open_weights = true
+
+[cost]
+input = 0.63
+output = 5.05
+
+[limit]
+context = 256000
+output = 4096
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/sap-ai-core/models/gemini-2.5-flash-lite.toml b/providers/sap-ai-core/models/gemini-2.5-flash-lite.toml
index 67a5ddda67..d5fece68cf 100644
--- a/providers/sap-ai-core/models/gemini-2.5-flash-lite.toml
+++ b/providers/sap-ai-core/models/gemini-2.5-flash-lite.toml
@@ -1,27 +1,21 @@
 name = "gemini-2.5-flash-lite"
-description = "Low-latency Gemini model for high-volume multimodal and agent workloads"
-family = "gemini-flash-lite"
-release_date = "2025-06-17"
+base_model = "google/gemini-2.5-flash-lite"
 last_updated = "2025-06-17"
 attachment = true
-reasoning = true
-# Vertex Gemini adapter: $.generationConfig.thinkingConfig.thinkingBudget is 0 (off), -1 (dynamic), or 512..24576. https://cloud.google.com/vertex-ai/generative-ai/docs/thinking (accessed 2026-06-25)
-reasoning_options = [{ type = "toggle" }, { type = "budget_tokens", min = 512, max = 24_576 }]
-temperature = true
-knowledge = "2025-01"
+# Vertex Gemini adapter: $.generationConfig.thinkingConfig.thinkingBudget is -1 (dynamic) or 512..24576; 0/off is unsupported. https://cloud.google.com/vertex-ai/generative-ai/docs/thinking (accessed 2026-06-25)
+reasoning_options = [{ type = "toggle" }, { type = "budget_tokens", min = 512, max = 24576 }]
 tool_call = true
 structured_output = true
 open_weights = false
 
 [cost]
-input = 0.10
-output = 0.40
+input = 0.08
+output = 0.28
 cache_read = 0.01
-input_audio = 0.30
 
 [limit]
-context = 1_048_576
-output = 65_536
+context = 1000000
+output = 65536
 
 [modalities]
 input = ["text", "image", "audio", "video", "pdf"]
diff --git a/providers/sap-ai-core/models/gemini-2.5-flash.toml b/providers/sap-ai-core/models/gemini-2.5-flash.toml
index 0b1c4ba865..4d113f3ab2 100644
--- a/providers/sap-ai-core/models/gemini-2.5-flash.toml
+++ b/providers/sap-ai-core/models/gemini-2.5-flash.toml
@@ -1,27 +1,21 @@
 name = "gemini-2.5-flash"
-description = "Fast Gemini model balancing multimodal reasoning, tool use, and cost"
-family = "gemini-flash"
-release_date = "2025-04-17"
+base_model = "google/gemini-2.5-flash"
 last_updated = "2025-06-05"
 attachment = true
-reasoning = true
 # Vertex Gemini adapter: $.generationConfig.thinkingConfig.thinkingBudget is 0 (off), -1 (dynamic), or 1..24576. https://cloud.google.com/vertex-ai/generative-ai/docs/thinking (accessed 2026-06-25)
-reasoning_options = [{ type = "toggle" }, { type = "budget_tokens", min = 0, max = 24_576 }]
-temperature = true
-knowledge = "2025-01"
+reasoning_options = [{ type = "toggle" }, { type = "budget_tokens", min = 0, max = 24576 }]
 tool_call = true
 structured_output = true
 open_weights = false
 
 [cost]
-input = 0.30
-output = 2.50
-cache_read = 0.030
-input_audio = 1.00
+input = 0.27
+output = 1.67
+cache_read = 0.03
 
 [limit]
-context = 1_048_576
-output = 65_536
+context = 1000000
+output = 65536
 
 [modalities]
 input = ["text", "image", "audio", "video", "pdf"]
diff --git a/providers/sap-ai-core/models/gemini-2.5-pro.toml b/providers/sap-ai-core/models/gemini-2.5-pro.toml
index c2da0ca58a..62eaf2075a 100644
--- a/providers/sap-ai-core/models/gemini-2.5-pro.toml
+++ b/providers/sap-ai-core/models/gemini-2.5-pro.toml
@@ -1,26 +1,26 @@
 name = "gemini-2.5-pro"
-description = "Advanced Gemini model for complex reasoning, coding, and multimodal analysis"
-family = "gemini-pro"
-release_date = "2025-03-25"
+base_model = "google/gemini-2.5-pro"
 last_updated = "2025-06-05"
 attachment = true
-reasoning = true
-# Vertex Gemini adapter: $.generationConfig.thinkingConfig.thinkingBudget is -1 (dynamic) or 128..32768; 0/off is unsupported. https://cloud.google.com/vertex-ai/generative-ai/docs/thinking (accessed 2026-06-25)
-reasoning_options = [{ type = "budget_tokens", min = 128, max = 32_768 }]
-temperature = true
-knowledge = "2025-01"
+# Vertex Gemini adapter: $.generationConfig.thinkingConfig.thinkingBudget is 0 (off), -1 (dynamic), or 128..32768. https://cloud.google.com/vertex-ai/generative-ai/docs/thinking (accessed 2026-06-25)
+reasoning_options = [{ type = "budget_tokens", min = 128, max = 32768 }]
 tool_call = true
 structured_output = true
 open_weights = false
 
 [cost]
-input = 1.25
-output = 10.00
-cache_read = 0.125
+input = 0.87
+output = 6.47
+cache_read = 0.09
+
+[[cost.tiers]]
+tier = { type = "context", size = 200000 }
+input = 1.67
+output = 9.66
 
 [limit]
-context = 1_048_576
-output = 65_536
+context = 1000000
+output = 65536
 
 [modalities]
 input = ["text", "image", "audio", "video", "pdf"]
diff --git a/providers/sap-ai-core/models/gemini-3.1-flash-lite.toml b/providers/sap-ai-core/models/gemini-3.1-flash-lite.toml
new file mode 100644
index 0000000000..2bc21972f2
--- /dev/null
+++ b/providers/sap-ai-core/models/gemini-3.1-flash-lite.toml
@@ -0,0 +1,18 @@
+name = "gemini-3.1-flash-lite"
+base_model = "google/gemini-3.1-flash-lite"
+last_updated = "2026-07-07"
+attachment = true
+# Vertex Gemini adapter: $.generationConfig.thinkingConfig.thinkingBudget is 0 (off), -1 (dynamic), or 1..24576. https://cloud.google.com/vertex-ai/generative-ai/docs/thinking (accessed 2026-06-25)
+reasoning_options = [{ type = "toggle" }, { type = "budget_tokens", min = 0, max = 24576 }]
+open_weights = false
+
+[cost]
+input = 0.2
+output = 1.08
+
+[limit]
+context = 1000000
+
+[modalities]
+input = ["text", "image"]
+output = ["text"]
diff --git a/providers/sap-ai-core/models/gpt-4.1-mini.toml b/providers/sap-ai-core/models/gpt-4.1-mini.toml
index 9bb3063755..734a003b28 100644
--- a/providers/sap-ai-core/models/gpt-4.1-mini.toml
+++ b/providers/sap-ai-core/models/gpt-4.1-mini.toml
@@ -1,24 +1,19 @@
 name = "gpt-4.1-mini"
-description = "Compact GPT model for low-latency assistance and high-volume workloads"
-family = "gpt-mini"
-release_date = "2025-04-14"
+base_model = "openai/gpt-4.1-mini"
 last_updated = "2025-04-14"
 attachment = true
-reasoning = false
-temperature = true
-knowledge = "2024-04"
 tool_call = true
 structured_output = true
 open_weights = false
 
 [cost]
-input = 0.40
-output = 1.60
-cache_read = 0.10
+input = 0.26
+output = 0.99
+cache_read = 0.07
 
 [limit]
-context = 1_047_576
-output = 32_768
+context = 1047576
+output = 32768
 
 [modalities]
 input = ["text", "image", "pdf"]
diff --git a/providers/sap-ai-core/models/gpt-4.1-nano.toml b/providers/sap-ai-core/models/gpt-4.1-nano.toml
new file mode 100644
index 0000000000..32d1303d6e
--- /dev/null
+++ b/providers/sap-ai-core/models/gpt-4.1-nano.toml
@@ -0,0 +1,16 @@
+name = "gpt-4.1-nano"
+base_model = "openai/gpt-4.1-nano"
+last_updated = "2026-07-07"
+attachment = true
+open_weights = false
+
+[cost]
+input = 0.08
+output = 0.26
+
+[limit]
+context = 1047576
+
+[modalities]
+input = ["text", "image"]
+output = ["text"]
diff --git a/providers/sap-ai-core/models/gpt-4.1.toml b/providers/sap-ai-core/models/gpt-4.1.toml
index e67e1573cc..733beb8d61 100644
--- a/providers/sap-ai-core/models/gpt-4.1.toml
+++ b/providers/sap-ai-core/models/gpt-4.1.toml
@@ -1,24 +1,19 @@
 name = "gpt-4.1"
-description = "GPT model for general reasoning, writing, coding, and tool-assisted tasks"
-family = "gpt"
-release_date = "2025-04-14"
+base_model = "openai/gpt-4.1"
 last_updated = "2025-04-14"
 attachment = true
-reasoning = false
-temperature = true
-knowledge = "2024-04"
 tool_call = true
 structured_output = true
 open_weights = false
 
 [cost]
-input = 2.00
-output = 8.00
-cache_read = 0.50
+input = 1.29
+output = 4.94
+cache_read = 0.32
 
 [limit]
-context = 1_047_576
-output = 32_768
+context = 1047576
+output = 32768
 
 [modalities]
 input = ["text", "image", "pdf"]
diff --git a/providers/sap-ai-core/models/gpt-5-mini.toml b/providers/sap-ai-core/models/gpt-5-mini.toml
index 6cf347be67..f6dce13347 100644
--- a/providers/sap-ai-core/models/gpt-5-mini.toml
+++ b/providers/sap-ai-core/models/gpt-5-mini.toml
@@ -1,26 +1,21 @@
 name = "gpt-5-mini"
-description = "Compact GPT model for low-latency assistance and high-volume workloads"
-family = "gpt-mini"
-release_date = "2025-08-07"
+base_model = "openai/gpt-5-mini"
 last_updated = "2025-08-07"
 attachment = true
-reasoning = true
+# Azure OpenAI deployment adapter: raw Chat uses top-level $.reasoning_effort = "minimal"|"low"|"medium"|"high"; there is no explicit off value for original gpt-5. https://learn.microsoft.com/en-us/azure/ai-foundry/openai/how-to/reasoning (accessed 2026-06-25)
 reasoning_options = [{ type = "effort", values = ["minimal", "low", "medium", "high"] }]
-temperature = false
-knowledge = "2024-05-30"
 tool_call = true
 structured_output = true
 open_weights = false
 
 [cost]
-input = 0.25
-output = 2.00
-cache_read = 0.025
+input = 0.24
+output = 1.41
+cache_read = 0.02
 
 [limit]
-context = 400_000
-input = 272_000
-output = 128_000
+context = 272000
+output = 128000
 
 [modalities]
 input = ["text", "image"]
diff --git a/providers/sap-ai-core/models/gpt-5-nano.toml b/providers/sap-ai-core/models/gpt-5-nano.toml
index 4c4e0014da..9da4d12cc2 100644
--- a/providers/sap-ai-core/models/gpt-5-nano.toml
+++ b/providers/sap-ai-core/models/gpt-5-nano.toml
@@ -1,26 +1,21 @@
 name = "gpt-5-nano"
-description = "Compact GPT model for low-latency assistance and high-volume workloads"
-family = "gpt-nano"
-release_date = "2025-08-07"
+base_model = "openai/gpt-5-nano"
 last_updated = "2025-08-07"
 attachment = true
-reasoning = true
+# Azure OpenAI deployment adapter: raw Chat uses top-level $.reasoning_effort = "minimal"|"low"|"medium"|"high"; there is no explicit off value for original gpt-5. https://learn.microsoft.com/en-us/azure/ai-foundry/openai/how-to/reasoning (accessed 2026-06-25)
 reasoning_options = [{ type = "effort", values = ["minimal", "low", "medium", "high"] }]
-temperature = false
-knowledge = "2024-05-30"
 tool_call = true
 structured_output = true
 open_weights = false
 
 [cost]
-input = 0.05
-output = 0.40
-cache_read = 0.005
+input = 0.11
+output = 0.34
+cache_read = 0.01
 
 [limit]
-context = 400_000
-input = 272_000
-output = 128_000
+context = 272000
+output = 128000
 
 [modalities]
 input = ["text", "image"]
diff --git a/providers/sap-ai-core/models/gpt-5.2.toml b/providers/sap-ai-core/models/gpt-5.2.toml
new file mode 100644
index 0000000000..7bf960c4ad
--- /dev/null
+++ b/providers/sap-ai-core/models/gpt-5.2.toml
@@ -0,0 +1,15 @@
+name = "gpt-5.2"
+base_model = "openai/gpt-5.2"
+last_updated = "2026-07-07"
+# Azure OpenAI deployment adapter: raw Chat uses top-level $.reasoning_effort = "none"|"minimal"|"low"|"medium"|"high"; "none" is off. https://learn.microsoft.com/en-us/azure/ai-foundry/openai/how-to/reasoning (accessed 2026-06-25)
+reasoning_options = [{ type = "effort", values = ["none", "minimal", "low", "medium", "high"] }]
+open_weights = false
+
+[cost]
+input = 1.25
+output = 9.44
+cache_read = 0.12
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/sap-ai-core/models/gpt-5.4.toml b/providers/sap-ai-core/models/gpt-5.4.toml
index a4bcd9b53d..d584967b6c 100644
--- a/providers/sap-ai-core/models/gpt-5.4.toml
+++ b/providers/sap-ai-core/models/gpt-5.4.toml
@@ -1,27 +1,21 @@
 name = "gpt-5.4"
-description = "Frontier GPT model for professional reasoning, coding, and multimodal work"
-family = "gpt"
-release_date = "2026-03-05"
+base_model = "openai/gpt-5.4"
 last_updated = "2026-03-05"
 attachment = true
-reasoning = true
 # Azure OpenAI deployment adapter: raw Chat uses top-level $.reasoning_effort = "none"|"low"|"medium"|"high"|"xhigh"; "none" is off. https://learn.microsoft.com/en-us/azure/ai-foundry/openai/how-to/reasoning (accessed 2026-06-25)
 reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh"] }]
-temperature = false
-knowledge = "2025-08-31"
 tool_call = true
 structured_output = true
 open_weights = false
 
 [cost]
-input = 2.50
-output = 15.00
-cache_read = 0.25
+input = 1.59
+output = 9.2
+cache_read = 0.16
 
 [limit]
-context = 1_050_000
-input = 922_000
-output = 128_000
+context = 1050000
+output = 128000
 
 [modalities]
 input = ["text", "image", "pdf"]
diff --git a/providers/sap-ai-core/models/gpt-5.5.toml b/providers/sap-ai-core/models/gpt-5.5.toml
index b6f360ae8c..f4a5da5be8 100644
--- a/providers/sap-ai-core/models/gpt-5.5.toml
+++ b/providers/sap-ai-core/models/gpt-5.5.toml
@@ -1,27 +1,21 @@
 name = "gpt-5.5"
-description = "Frontier GPT model for professional reasoning, coding, and multimodal work"
-family = "gpt"
-release_date = "2026-04-23"
+base_model = "openai/gpt-5.5"
 last_updated = "2026-04-23"
 attachment = true
-reasoning = true
 # Azure OpenAI deployment adapter: raw Chat uses top-level $.reasoning_effort = "none"|"low"|"medium"|"high"|"xhigh"; "none" is off. https://learn.microsoft.com/en-us/azure/ai-foundry/openai/how-to/reasoning (accessed 2026-06-25)
 reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh"] }]
-temperature = false
-knowledge = "2025-12-01"
 tool_call = true
 structured_output = true
 open_weights = false
 
 [cost]
-input = 5.00
-output = 30.00
-cache_read = 0.50
+input = 3.42
+output = 20.15
+cache_read = 0.34
 
 [limit]
-context = 1_050_000
-input = 922_000
-output = 128_000
+context = 1050000
+output = 128000
 
 [modalities]
 input = ["text", "image", "pdf"]
diff --git a/providers/sap-ai-core/models/gpt-5.toml b/providers/sap-ai-core/models/gpt-5.toml
index 7406a033ad..700f047b26 100644
--- a/providers/sap-ai-core/models/gpt-5.toml
+++ b/providers/sap-ai-core/models/gpt-5.toml
@@ -1,27 +1,21 @@
 name = "gpt-5"
-description = "GPT model for general reasoning, writing, coding, and tool-assisted tasks"
-family = "gpt"
-release_date = "2025-08-07"
+base_model = "openai/gpt-5"
 last_updated = "2025-08-07"
 attachment = true
-reasoning = true
 # Azure OpenAI deployment adapter: raw Chat uses top-level $.reasoning_effort = "minimal"|"low"|"medium"|"high"; there is no explicit off value for original gpt-5. https://learn.microsoft.com/en-us/azure/ai-foundry/openai/how-to/reasoning (accessed 2026-06-25)
 reasoning_options = [{ type = "effort", values = ["minimal", "low", "medium", "high"] }]
-temperature = false
-knowledge = "2024-09-30"
 tool_call = true
 structured_output = true
 open_weights = false
 
 [cost]
-input = 1.25
-output = 10.00
-cache_read = 0.125
+input = 0.91
+output = 6.77
+cache_read = 0.09
 
 [limit]
-context = 400_000
-input = 272_000
-output = 128_000
+context = 272000
+output = 128000
 
 [modalities]
 input = ["text", "image"]
diff --git a/providers/sap-ai-core/models/mistralai--mistral-large-instruct.toml b/providers/sap-ai-core/models/mistralai--mistral-large-instruct.toml
new file mode 100644
index 0000000000..c0cf653434
--- /dev/null
+++ b/providers/sap-ai-core/models/mistralai--mistral-large-instruct.toml
@@ -0,0 +1,21 @@
+name = "mistralai--mistral-large-instruct"
+description = "Mistral Large 2"
+release_date = "2026-07-07"
+last_updated = "2026-07-07"
+attachment = false
+reasoning = false
+temperature = true
+tool_call = true
+open_weights = true
+
+[cost]
+input = 1.12
+output = 3.2
+
+[limit]
+context = 64000
+output = 4096
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/sap-ai-core/models/mistralai--mistral-medium-instruct.toml b/providers/sap-ai-core/models/mistralai--mistral-medium-instruct.toml
new file mode 100644
index 0000000000..2e9afffc3a
--- /dev/null
+++ b/providers/sap-ai-core/models/mistralai--mistral-medium-instruct.toml
@@ -0,0 +1,21 @@
+name = "mistralai--mistral-medium-instruct"
+description = "Mistral Medium 3 Instruct"
+release_date = "2026-07-07"
+last_updated = "2026-07-07"
+attachment = true
+reasoning = false
+temperature = true
+tool_call = true
+open_weights = true
+
+[cost]
+input = 0.36
+output = 1.22
+
+[limit]
+context = 128000
+output = 4096
+
+[modalities]
+input = ["text", "image"]
+output = ["text"]
diff --git a/providers/sap-ai-core/models/mistralai--mistral-small.toml b/providers/sap-ai-core/models/mistralai--mistral-small.toml
new file mode 100644
index 0000000000..670d5ad20a
--- /dev/null
+++ b/providers/sap-ai-core/models/mistralai--mistral-small.toml
@@ -0,0 +1,16 @@
+name = "mistralai--mistral-small"
+base_model = "mistral/mistral-small-2506"
+last_updated = "2026-07-07"
+attachment = true
+open_weights = true
+
+[cost]
+input = 0.07
+output = 0.28
+
+[limit]
+context = 128000
+
+[modalities]
+input = ["text", "image"]
+output = ["text"]
diff --git a/providers/sap-ai-core/models/nvidia--llama-3.2-nv-embedqa-1b.toml b/providers/sap-ai-core/models/nvidia--llama-3.2-nv-embedqa-1b.toml
new file mode 100644
index 0000000000..8eb413aa0d
--- /dev/null
+++ b/providers/sap-ai-core/models/nvidia--llama-3.2-nv-embedqa-1b.toml
@@ -0,0 +1,21 @@
+name = "nvidia--llama-3.2-nv-embedqa-1b"
+description = "Nvidia llama-3.2-nv-embedqa-1b Embedding model"
+release_date = "2026-07-07"
+last_updated = "2026-07-07"
+attachment = false
+reasoning = false
+temperature = false
+tool_call = false
+open_weights = false
+
+[cost]
+input = 0.07
+output = 0
+
+[limit]
+context = 131072
+output = 4096
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/sap-ai-core/models/sap-abap-1.toml b/providers/sap-ai-core/models/sap-abap-1.toml
new file mode 100644
index 0000000000..aa99a7e8eb
--- /dev/null
+++ b/providers/sap-ai-core/models/sap-abap-1.toml
@@ -0,0 +1,21 @@
+name = "sap-abap-1"
+description = "SAP ABAP 1"
+release_date = "2026-07-07"
+last_updated = "2026-07-07"
+attachment = false
+reasoning = false
+temperature = true
+tool_call = true
+open_weights = false
+
+[cost]
+input = 0.48
+output = 1.7
+
+[limit]
+context = 32768
+output = 4096
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/sap-ai-core/models/sonar-pro.toml b/providers/sap-ai-core/models/sonar-pro.toml
index 9a9e70d191..ff1a17a382 100644
--- a/providers/sap-ai-core/models/sonar-pro.toml
+++ b/providers/sap-ai-core/models/sonar-pro.toml
@@ -1,22 +1,17 @@
 name = "sonar-pro"
-description = "Advanced Sonar search model for deeper research and cited synthesis"
-family = "sonar-pro"
-release_date = "2024-01-01"
+base_model = "perplexity/sonar-pro"
 last_updated = "2025-09-01"
 attachment = true
-reasoning = false
-temperature = true
 tool_call = false
-knowledge = "2025-09-01"
 open_weights = false
 
 [cost]
-input = 3.00
-output = 15.00
+input = 2.43
+output = 11.85
 
 [limit]
-context = 200_000
-output = 8_192
+context = 200000
+output = 8192
 
 [modalities]
 input = ["text", "image"]
diff --git a/providers/sap-ai-core/models/sonar.toml b/providers/sap-ai-core/models/sonar.toml
index 9d8859641a..c5269d8c29 100644
--- a/providers/sap-ai-core/models/sonar.toml
+++ b/providers/sap-ai-core/models/sonar.toml
@@ -1,23 +1,18 @@
 name = "sonar"
-description = "Sonar search model for current answers, retrieval, and citation-backed chat"
-family = "sonar"
-release_date = "2024-01-01"
+base_model = "perplexity/sonar"
 last_updated = "2025-09-01"
 attachment = false
-reasoning = false
-temperature = true
 tool_call = false
-knowledge = "2025-09-01"
 open_weights = false
 
 [cost]
-input = 1.00
-output = 1.00
+input = 0.86
+output = 0.86
 
 [limit]
-context = 128_000
-output = 4_096
+context = 128000
+output = 4096
 
 [modalities]
-input = ["text"]
+input = ["text", "image"]
 output = ["text"]
diff --git a/providers/sap-ai-core/models/text-embedding-3-large.toml b/providers/sap-ai-core/models/text-embedding-3-large.toml
new file mode 100644
index 0000000000..af9ddca41b
--- /dev/null
+++ b/providers/sap-ai-core/models/text-embedding-3-large.toml
@@ -0,0 +1,21 @@
+name = "text-embedding-3-large"
+description = "Azure OpenAI Text Embedding 3 Large Model"
+release_date = "2026-07-07"
+last_updated = "2026-07-07"
+attachment = false
+reasoning = false
+temperature = false
+tool_call = false
+open_weights = false
+
+[cost]
+input = 0.09
+output = 0
+
+[limit]
+context = 131072
+output = 4096
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/sap-ai-core/models/text-embedding-3-small.toml b/providers/sap-ai-core/models/text-embedding-3-small.toml
new file mode 100644
index 0000000000..c47cf4affc
--- /dev/null
+++ b/providers/sap-ai-core/models/text-embedding-3-small.toml
@@ -0,0 +1,21 @@
+name = "text-embedding-3-small"
+description = "Azure OpenAI Text Embedding 3 Small Model"
+release_date = "2026-07-07"
+last_updated = "2026-07-07"
+attachment = false
+reasoning = false
+temperature = false
+tool_call = false
+open_weights = false
+
+[cost]
+input = 0.02
+output = 0
+
+[limit]
+context = 131072
+output = 4096
+
+[modalities]
+input = ["text"]
+output = ["text"]