feat: add Nebius Token Factory models

providers/nebius/models/BAAI/bge-en-icl.toml

@@ -0,0 +1,24 @@
+name = "BGE-ICL"
+family = "text-embedding"
+attachment = false
+reasoning = false
+tool_call = false
+structured_output = false
+temperature = false
+knowledge = "2024-06"
+release_date = "2024-07-30"
+last_updated = "2026-02-04"
+open_weights = true
+
+[cost]
+input = 0.01
+output = 0.00
+
+[limit]
+context = 32_768
+input = 32_768
+output = 0
+
+[modalities]
+input = ["text"]
+output = ["text"]

providers/nebius/models/BAAI/bge-multilingual-gemma2.toml

@@ -0,0 +1,24 @@
+name = "bge-multilingual-gemma2"
+family = "text-embedding"
+attachment = false
+reasoning = false
+tool_call = false
+structured_output = false
+temperature = false
+knowledge = "2024-06"
+release_date = "2024-07-30"
+last_updated = "2026-02-04"
+open_weights = true
+
+[cost]
+input = 0.01
+output = 0.00
+
+[limit]
+context = 8_192
+input = 8_192
+output = 0
+
+[modalities]
+input = ["text"]
+output = ["text"]

providers/nebius/models/MiniMaxAI/minimax-m2.1.toml

@@ -0,0 +1,29 @@
+name = "MiniMax-M2.1"
+attachment = false
+reasoning = true
+tool_call = true
+structured_output = true
+temperature = true
+knowledge = "2025-10"
+release_date = "2026-02-01"
+last_updated = "2026-02-04"
+open_weights = true
+
+[cost]
+input = 0.30
+output = 1.20
+reasoning = 1.20
+cache_read = 0.03
+cache_write = 0.375
+
+[limit]
+context = 128_000
+input = 120_000
+output = 8_192
+
+[modalities]
+input = ["text"]
+output = ["text"]
+
+[interleaved]
+field = "reasoning_content"

providers/nebius/models/NousResearch/hermes-4-405b.toml

@@ -1,22 +1,29 @@
-name = "Hermes-4 405B"
-family = "hermes"
-release_date = "2024-08-01"
-last_updated = "2025-10-04"
+name = "Hermes-4-405B"
 attachment = false
 reasoning = true
-temperature = true
-knowledge = "2024-07"
 tool_call = true
-open_weights = false
+structured_output = true
+temperature = true
+knowledge = "2025-11"
+release_date = "2026-01-30"
+last_updated = "2026-02-04"
+open_weights = true
 
 [cost]
 input = 1.00
 output = 3.00
+reasoning = 3.00
+cache_read = 0.10
+cache_write = 1.25
 
 [limit]
-context = 131072
-output = 8192
+context = 128_000
+input = 120_000
+output = 8_192
 
 [modalities]
 input = ["text"]
 output = ["text"]
+
+[interleaved]
+field = "reasoning_content"

providers/nebius/models/NousResearch/hermes-4-70b.toml

@@ -1,22 +1,29 @@
-name = "Hermes 4 70B"
-family = "hermes"
-release_date = "2024-08-01"
-last_updated = "2025-10-04"
+name = "Hermes-4-70B"
 attachment = false
 reasoning = true
-temperature = true
-knowledge = "2024-07"
 tool_call = true
-open_weights = false
+structured_output = true
+temperature = true
+knowledge = "2025-11"
+release_date = "2026-01-30"
+last_updated = "2026-02-04"
+open_weights = true
 
 [cost]
 input = 0.13
 output = 0.40
+reasoning = 0.40
+cache_read = 0.013
+cache_write = 0.16
 
 [limit]
-context = 131072
-output = 8192
+context = 128_000
+input = 120_000
+output = 8_192
 
 [modalities]
 input = ["text"]
 output = ["text"]
+
+[interleaved]
+field = "reasoning_content"

providers/nebius/models/PrimeIntellect/intellect-3.toml

@@ -0,0 +1,25 @@
+name = "INTELLECT-3"
+attachment = false
+reasoning = false
+tool_call = true
+structured_output = true
+temperature = true
+knowledge = "2025-10"
+release_date = "2026-01-25"
+last_updated = "2026-02-04"
+open_weights = true
+
+[cost]
+input = 0.20
+output = 1.10
+cache_read = 0.02
+cache_write = 0.25
+
+[limit]
+context = 128_000
+input = 120_000
+output = 8_192
+
+[modalities]
+input = ["text"]
+output = ["text"]

providers/nebius/models/black-forest-labs/flux-dev.toml

@@ -0,0 +1,23 @@
+name = "FLUX.1-dev"
+attachment = false
+reasoning = false
+tool_call = false
+structured_output = false
+temperature = false
+knowledge = "2024-07"
+release_date = "2024-08-01"
+last_updated = "2026-02-04"
+open_weights = true
+
+[cost]
+input = 0.00
+output = 0.00
+
+[limit]
+context = 77
+input = 77
+output = 0
+
+[modalities]
+input = ["text"]
+output = ["image"]

providers/nebius/models/black-forest-labs/flux-schnell.toml

@@ -0,0 +1,23 @@
+name = "FLUX.1-schnell"
+attachment = false
+reasoning = false
+tool_call = false
+structured_output = false
+temperature = false
+knowledge = "2024-07"
+release_date = "2024-08-01"
+last_updated = "2026-02-04"
+open_weights = true
+
+[cost]
+input = 0.00
+output = 0.00
+
+[limit]
+context = 77
+input = 77
+output = 0
+
+[modalities]
+input = ["text"]
+output = ["image"]

providers/nebius/models/deepseek-ai/deepseek-r1-0528-fast.toml

@@ -0,0 +1,22 @@
+name = "DeepSeek R1 0528 Fast"
+family = "deepseek"
+release_date = "2025-01-01"
+last_updated = "2025-02-04"
+attachment = false
+reasoning = true
+temperature = true
+knowledge = "2025-01"
+tool_call = true
+open_weights = true
+
+[cost]
+input = 2.00
+output = 6.00
+
+[limit]
+context = 131072
+output = 8192
+
+[modalities]
+input = ["text"]
+output = ["text"]

providers/nebius/models/deepseek-ai/deepseek-r1-0528.toml

@@ -0,0 +1,29 @@
+name = "DeepSeek-R1-0528"
+attachment = false
+reasoning = true
+tool_call = true
+structured_output = true
+temperature = true
+knowledge = "2025-11"
+release_date = "2026-01-15"
+last_updated = "2026-02-04"
+open_weights = true
+
+[cost]
+input = 0.80
+output = 2.40
+reasoning = 2.40
+cache_read = 0.08
+cache_write = 1.00
+
+[limit]
+context = 128_000
+input = 120_000
+output = 32_768
+
+[modalities]
+input = ["text"]
+output = ["text"]
+
+[interleaved]
+field = "reasoning_content"

providers/nebius/models/deepseek-ai/deepseek-v3-0324-fast.toml

@@ -0,0 +1,25 @@
+name = "DeepSeek-V3-0324 (Fast)"
+attachment = false
+reasoning = false
+tool_call = true
+structured_output = true
+temperature = true
+knowledge = "2024-12"
+release_date = "2025-03-24"
+last_updated = "2026-02-04"
+open_weights = true
+
+[cost]
+input = 0.75
+output = 2.25
+cache_read = 0.075
+cache_write = 0.28125
+
+[limit]
+context = 128_000
+input = 120_000
+output = 8_192
+
+[modalities]
+input = ["text"]
+output = ["text"]

providers/nebius/models/deepseek-ai/deepseek-v3-0324.toml

@@ -0,0 +1,25 @@
+name = "DeepSeek-V3-0324"
+attachment = false
+reasoning = false
+tool_call = true
+structured_output = true
+temperature = true
+knowledge = "2024-12"
+release_date = "2025-03-24"
+last_updated = "2026-02-04"
+open_weights = true
+
+[cost]
+input = 0.50
+output = 1.50
+cache_read = 0.05
+cache_write = 0.1875
+
+[limit]
+context = 128_000
+input = 120_000
+output = 8_192
+
+[modalities]
+input = ["text"]
+output = ["text"]

providers/nebius/models/deepseek-ai/deepseek-v3.2.toml

@@ -0,0 +1,29 @@
+name = "DeepSeek-V3.2"
+attachment = false
+reasoning = true
+tool_call = true
+structured_output = true
+temperature = true
+knowledge = "2025-11"
+release_date = "2026-01-20"
+last_updated = "2026-02-04"
+open_weights = true
+
+[cost]
+input = 0.30
+output = 0.45
+reasoning = 0.45
+cache_read = 0.03
+cache_write = 0.375
+
+[limit]
+context = 128_000
+input = 120_000
+output = 8_192
+
+[modalities]
+input = ["text"]
+output = ["text"]
+
+[interleaved]
+field = "reasoning_content"

providers/nebius/models/deepseek-ai/deepseek-v3.toml

@@ -1,13 +1,14 @@
 name = "DeepSeek V3"
 family = "deepseek"
 release_date = "2024-05-07"
-last_updated = "2025-10-04"
+last_updated = "2026-02-04"
 attachment = false
 reasoning = true
 temperature = true
 knowledge = "2024-04"
 tool_call = true
 open_weights = false
+status = "deprecated"
 
 [cost]
 input = 0.50