diff --git a/providers/sambanova/models/DeepSeek-V3.1.toml b/providers/sambanova/models/DeepSeek-V3.1.toml
new file mode 100644
index 000000000..d5edba297
--- /dev/null
+++ b/providers/sambanova/models/DeepSeek-V3.1.toml
@@ -0,0 +1,24 @@
+name = "DeepSeek-V3.1"
+family = "deepseek"
+release_date = "2025-08-21"
+last_updated = "2026-06-10"
+attachment = false
+reasoning = true
+reasoning_options = [{ type = "toggle" }]
+temperature = true
+knowledge = "2024-07"
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 3.00
+output = 4.50
+
+[limit]
+context = 131_072
+output = 7_168
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/sambanova/models/DeepSeek-V3.2.toml b/providers/sambanova/models/DeepSeek-V3.2.toml
new file mode 100644
index 000000000..badf69244
--- /dev/null
+++ b/providers/sambanova/models/DeepSeek-V3.2.toml
@@ -0,0 +1,25 @@
+name = "DeepSeek-V3.2"
+family = "deepseek"
+release_date = "2025-12-01"
+last_updated = "2026-06-10"
+attachment = false
+reasoning = true
+reasoning_options = [{ type = "toggle" }]
+temperature = true
+knowledge = "2024-12"
+tool_call = true
+structured_output = true
+open_weights = true
+status = "beta"
+
+[cost]
+input = 3.00
+output = 4.50
+
+[limit]
+context = 32_768
+output = 7_168
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/sambanova/models/Meta-Llama-3.3-70B-Instruct.toml b/providers/sambanova/models/Meta-Llama-3.3-70B-Instruct.toml
new file mode 100644
index 000000000..134595712
--- /dev/null
+++ b/providers/sambanova/models/Meta-Llama-3.3-70B-Instruct.toml
@@ -0,0 +1,11 @@
+base_model = "meta/llama-3.3-70b-instruct"
+attachment = false
+structured_output = true
+
+[cost]
+input = 0.60
+output = 1.20
+
+[limit]
+context = 131_072
+output = 3_072
diff --git a/providers/sambanova/models/MiniMax-M2.7.toml b/providers/sambanova/models/MiniMax-M2.7.toml
new file mode 100644
index 000000000..50e870465
--- /dev/null
+++ b/providers/sambanova/models/MiniMax-M2.7.toml
@@ -0,0 +1,10 @@
+base_model = "minimax/MiniMax-M2.7"
+reasoning = false
+
+[cost]
+input = 0.60
+output = 2.40
+
+[limit]
+context = 196_608
+output = 196_608
diff --git a/providers/sambanova/models/gemma-4-31B-it.toml b/providers/sambanova/models/gemma-4-31B-it.toml
new file mode 100644
index 000000000..419ac4695
--- /dev/null
+++ b/providers/sambanova/models/gemma-4-31B-it.toml
@@ -0,0 +1,18 @@
+base_model = "google/gemma-4-31b-it"
+attachment = false
+reasoning = false
+tool_call = false
+structured_output = false
+status = "beta"
+
+[cost]
+input = 0.38
+output = 1.15
+
+[limit]
+context = 131_072
+output = 131_072
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/sambanova/models/gpt-oss-120b.toml b/providers/sambanova/models/gpt-oss-120b.toml
new file mode 100644
index 000000000..456a2055d
--- /dev/null
+++ b/providers/sambanova/models/gpt-oss-120b.toml
@@ -0,0 +1,24 @@
+name = "GPT OSS 120B"
+family = "gpt-oss"
+release_date = "2025-08-05"
+last_updated = "2026-06-10"
+attachment = false
+reasoning = true
+reasoning_options = [{ type = "effort", values = ["low", "medium", "high"] }]
+temperature = true
+knowledge = "2024-06"
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.22
+output = 0.59
+
+[limit]
+context = 131_072
+output = 131_072
+
+[modalities]
+input = ["text"]
+output = ["text"]
diff --git a/providers/sambanova/provider.toml b/providers/sambanova/provider.toml
new file mode 100644
index 000000000..db585a9cb
--- /dev/null
+++ b/providers/sambanova/provider.toml
@@ -0,0 +1,5 @@
+name = "SambaNova"
+env = ["SAMBANOVA_API_KEY"]
+npm = "@ai-sdk/openai-compatible"
+doc = "https://docs.sambanova.ai/docs/en/models/sambacloud-models"
+api = "https://api.sambanova.ai/v1"