chore(sync): update Inceptron model catalog (#5365 )

Co-authored-by: opencode-agent[bot] <opencode-agent[bot]@users.noreply.github.com>
chore(sync): update NanoGPT model catalog (#5364 )
2026-08-24 11:26:00 +00:00 · 2026-08-24 10:28:07 +00:00 · 2026-08-24 18:16:36 +08:00 · 2026-08-24 09:32:07 +00:00 · 2026-08-24 09:31:17 +00:00 · 2026-08-24 07:39:54 +00:00
136 changed files with 1046 additions and 499 deletions
@@ -1,14 +1,16 @@
+# https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813
 name = "DeepSeek V4 Pro 0813"
 description = "DeepSeek V4 Pro snapshot with million-token context and support for thinking and non-thinking modes"
 family = "deepseek-thinking"
 release_date = "2026-08-12"
-last_updated = "2026-08-12"
+last_updated = "2026-08-22"
 attachment = false
 reasoning = true
 temperature = true
 tool_call = true
 structured_output = true
-open_weights = false
+open_weights = true
+license = "MIT"

 [limit]
 context = 1_000_000
@@ -17,3 +19,7 @@ output = 384_000
 [modalities]
 input = ["text"]
 output = ["text"]
+
+[[weights]]
+label = "Hugging Face"
+url = "https://huggingface.co/deepseek-ai/DeepSeek-V4-Pro-0813"
@@ -68,7 +68,10 @@ export interface CostTier extends Cost {

 /** Pricing for a provider's model, including context-size tiers. */
 export interface ModelCost extends Cost {
-  /** Legacy compatibility field: pricing applied beyond 200K context. Prefer `tiers`. */
+  /**
+   * Legacy compatibility field for context-tier pricing.
+   * @deprecated Use `tiers` to read the exact context threshold.
+   */
  context_over_200k?: Cost
  /** Context-size-based pricing tiers. */
  tiers?: CostTier[]
@@ -0,0 +1,4 @@
+<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 24 24" fill="currentColor" role="img">
+  <title>AgentRouter</title>
+  <path d="M11.8 5C7.8 4.2 3 9.6 4.2 15.6c.4 3 3 4.6 6.6 3-1.8-3-2.6-7.8 1-13.6Zm.4 14c4 .8 8.8-4.6 7.6-10.6-.4-3-3-4.6-6.6-3 1.8 3 2.6 7.8-1 13.6ZM12 9.5c.2 1.5 1 2.3 2.5 2.5-1.5.2-2.3 1-2.5 2.5-.2-1.5-1-2.3-2.5-2.5 1.5-.2 2.3-1 2.5-2.5Z"/>
+</svg>
@@ -0,0 +1,8 @@
+# Catalog and endpoint support: https://agentrouter.org/api/pricing
+# AgentRouter publishes relative token ratios, not USD prices, so cost is intentionally omitted.
+base_model = "anthropic/claude-opus-4-8"
+reasoning_options = [{ type = "effort", values = ["low", "medium", "high", "xhigh", "max"] }]
+
+[provider]
+npm = "@ai-sdk/anthropic"
+api = "https://agentrouter.org/v1"
@@ -0,0 +1,8 @@
+# Catalog and endpoint support: https://agentrouter.org/api/pricing
+# AgentRouter publishes relative token ratios, not USD prices, so cost is intentionally omitted.
+base_model = "anthropic/claude-opus-5"
+reasoning_options = [{ type = "effort", values = ["low", "medium", "high", "xhigh", "max"] }]
+
+[provider]
+npm = "@ai-sdk/anthropic"
+api = "https://agentrouter.org/v1"
@@ -0,0 +1,4 @@
+# Catalog and endpoint support: https://agentrouter.org/api/pricing
+# AgentRouter publishes relative token ratios, not USD prices, so cost is intentionally omitted.
+base_model = "openai/gpt-5.6-sol"
+reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }]
@@ -0,0 +1,5 @@
+name = "AgentRouter"
+env = ["AGENTROUTER_API_KEY"]
+npm = "@ai-sdk/openai-compatible"
+api = "https://agentrouter.org/v1"
+doc = "https://agentrouter.org/docs/opencode.html"
@@ -0,0 +1,13 @@
+# attachment are not supported on AKI.IO for security reasons and therefore are disabled
+base_model = "deepseek/deepseek-v4-flash-0731"
+reasoning_options = [{ type = "effort", values = ["none", "low", "high", "max"] }]
+
+# cost as described on the AKI.IO pricing page at https://aki.io/#pricing
+[cost]
+input = 0.20
+output = 0.50
+cache_read = 0.10
+
+[limit]
+context = 1_048_576
+output = 81_920
@@ -2,7 +2,6 @@
 # AKI.IO uses the reasoning_option 'none' to disable reasoning
 base_model = "google/gemma-4-26b-a4b-it"
 attachment = false
-reasoning = true
 reasoning_options = [{ type = "effort", values = ["none", "high"] }]

 [cost]
@@ -4,9 +4,11 @@ base_model = "moonshotai/kimi-k2.7-code"
 attachment = false
 reasoning_options = []

+# cost as described on the AKI.IO pricing page at https://aki.io/#pricing
 [cost]
 input = 0.86
 output = 3.00
+cache_read = 0.18

 [limit]
 context = 262_144
@@ -1,12 +0,0 @@
-# attachment are not supported on AKI.IO for security reasons and therefore are disabled
-base_model = "minimax/MiniMax-M2.5"
-attachment = false 
-reasoning_options = []
-
-[cost]
-input = 0.25
-output = 1.20
-
-[limit]
-context = 196_608
-output = 65_536
@@ -0,0 +1,14 @@
+# attachment are not supported on AKI.IO for security reasons and therefore are disabled
+# AKI.IO uses the reasoning_option 'none' to disable reasoning
+base_model = "alibaba/qwen3.8-27b"
+attachment = false
+reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "xhigh"] }]
+
+# cost as described on the AKI.IO pricing page at https://aki.io/#pricing
+[cost]
+input = 0.30
+output = 2.20
+cache_read = 0.10
+
+[modalities]
+input = ["text", "image"]
@@ -21,6 +21,8 @@ max = 81_920
 input = 0.115
 output = 0.287
 reasoning = 1.147
+cache_read = 0.012
+cache_write = 0.144

 [limit]
 context = 1_000_000
@@ -6,9 +6,9 @@ last_updated = "2026-08-16"
 reasoning_options = []

 [cost]
-input = 0.4
-output = 3
-cache_read = 0.039999999999999994
+input = 0.35
+output = 2.75
+cache_read = 0.03499999999999999

 [limit]
 output = 65_536
@@ -1,16 +0,0 @@
-base_model = "zhipuai/glm-4.7-flash"
-# Crof's request docs establish no model-specific toggle, effort, or budget.
-# https://crof.ai/docs.md (accessed 2026-06-25)
-reasoning_options = []
-
-[cost]
-input = 0.04
-output = 0.3
-cache_read = 0.008
-cache_write = 0
-
-[limit]
-context = 202_752
-
-[provider]
-npm = "@ai-sdk/openai-compatible"
@@ -1,20 +0,0 @@
-base_model = "zhipuai/glm-4.7"
-# Crof's request docs establish no model-specific toggle, effort, or budget.
-# https://crof.ai/docs.md (accessed 2026-06-25)
-reasoning_options = []
-
-[interleaved]
-field = "reasoning_content"
-
-[cost]
-input = 0.25
-output = 1.1
-cache_read = 0.05
-cache_write = 0
-
-[limit]
-context = 202_752
-output = 202_752
-
-[provider]
-npm = "@ai-sdk/openai-compatible"
@@ -1,20 +0,0 @@
-base_model = "zhipuai/glm-5"
-# Crof's request docs establish no model-specific toggle, effort, or budget.
-# https://crof.ai/docs.md (accessed 2026-06-25)
-reasoning_options = []
-
-[interleaved]
-field = "reasoning_content"
-
-[cost]
-input = 0.48
-output = 1.9
-cache_read = 0.1
-cache_write = 0
-
-[limit]
-context = 202_752
-output = 202_752
-
-[provider]
-npm = "@ai-sdk/openai-compatible"
@@ -1,28 +0,0 @@
-name = "Kimi K2.5 (Lightning)"
-description = "Kimi multimodal agent model for visual understanding, coding, and planning"
-family = "kimi-k2"
-release_date = "2026-02-06"
-last_updated = "2026-02-06"
-attachment = true
-reasoning = true
-reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high"] }]
-temperature = false
-tool_call = true
-structured_output = true
-open_weights = true
-
-[interleaved]
-field = "reasoning_content"
-
-[cost]
-input = 1.00
-output = 3.00
-cache_read = 0.20
-
-[limit]
-context = 131_072
-output = 32_768
-
-[modalities]
-input = ["text", "image", "video"]
-output = ["text"]
@@ -1,14 +0,0 @@
-base_model = "minimax/MiniMax-M2.5"
-# Crof's request docs establish no model-specific toggle, effort, or budget.
-# https://crof.ai/docs.md (accessed 2026-06-25)
-reasoning = true
-reasoning_options = []
-
-[cost]
-input = 0.11
-output = 0.95
-cache_read = 0.02
-cache_write = 0.375
-
-[provider]
-npm = "@ai-sdk/openai-compatible"
@@ -1,13 +1,13 @@
-base_model = "moonshotai/kimi-k2.5"
+base_model = "alibaba/qwen3.8-27b"
 reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high"] }]

-[interleaved]
-field = "reasoning_content"
-
 [cost]
-input = 0.35
-output = 1.7
-cache_read = 0.07
+input = 0.25
+output = 2.1
+cache_read = 0.06
+
+[limit]
+output = 262_144

 [provider]
 npm = "@ai-sdk/openai-compatible"
@@ -1,6 +1,4 @@
 base_model = "deepseek/deepseek-v4-pro-0813"
-tool_call = false
-structured_output = false

 [[reasoning_options]]
 type = "effort"
@@ -5,5 +5,5 @@ type = "effort"
 values = ["none", "low", "high", "max"]

 [cost]
-input = 0.176
-output = 0.528
+input = 0.352
+output = 1.056
@@ -5,5 +5,5 @@ type = "effort"
 values = ["none", "low", "high", "max"]

 [cost]
-input = 0.627
-output = 1.881
+input = 1.1616
+output = 3.4848
@@ -3,6 +3,15 @@ base_model = "deepseek/deepseek-v4-flash-0731"
 [[reasoning_options]]
 type = "toggle"

+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+min = 1
+max = 393_216
+
 [cost]
 input = 0.424
 output = 1.272
@@ -4,6 +4,10 @@ last_updated = "2026-06-12"
 [[reasoning_options]]
 type = "toggle"

+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
 [[reasoning_options]]
 type = "budget_tokens"
 min = 1
@@ -4,6 +4,10 @@ last_updated = "2026-06-12"
 [[reasoning_options]]
 type = "toggle"

+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
 [[reasoning_options]]
 type = "budget_tokens"
 min = 1
@@ -5,6 +5,10 @@ last_updated = "2026-06-12"
 [[reasoning_options]]
 type = "toggle"

+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
 [[reasoning_options]]
 type = "budget_tokens"
 min = 1
@@ -4,6 +4,10 @@ last_updated = "2026-06-12"
 [[reasoning_options]]
 type = "toggle"

+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high", "max"]
+
 [cost]
 input = 0.8939
 output = 3.7131
@@ -8,7 +8,7 @@ field = "reasoning_content"
 [cost]
 input = 0.67
 output = 3.4
-cache_read = 0.17
+cache_read = 0.19
 cache_write = 0

 [modalities]
@@ -12,3 +12,4 @@ cache_read = 0.028

 [limit]
 context = 1_048_576
+output = 131_072
@@ -6,8 +6,8 @@ type = "effort"
 values = ["none", "high"]

 [cost]
-input = 0.05
-output = 0.25
+input = 0.042
+output = 0.22

 [limit]
 output = 16_384
@@ -1,24 +0,0 @@
-name = "inclusionAI: Ling-2.6-1T (retires Aug 24)"
-description = "Tool-capable chat model for instruction following and agentic application workflows"
-family = "ling"
-release_date = "2026-04-23"
-last_updated = "2026-04-23"
-attachment = false
-reasoning = false
-temperature = true
-tool_call = true
-structured_output = true
-open_weights = false
-
-[cost]
-input = 0.3
-output = 2.5
-cache_read = 0.06
-
-[limit]
-context = 262_144
-output = 32_768
-
-[modalities]
-input = ["text"]
-output = ["text"]
@@ -1,24 +0,0 @@
-name = "inclusionAI: Ling-2.6-flash (retires Aug 24)"
-description = "Efficient model for low-latency assistance, extraction, and routine automation"
-family = "ling"
-release_date = "2026-04-21"
-last_updated = "2026-04-21"
-attachment = false
-reasoning = false
-temperature = true
-tool_call = true
-structured_output = true
-open_weights = false
-
-[cost]
-input = 0.1
-output = 0.3
-cache_read = 0.02
-
-[limit]
-context = 262_144
-output = 32_768
-
-[modalities]
-input = ["text"]
-output = ["text"]
@@ -1,28 +0,0 @@
-name = "inclusionAI: Ring-2.6-1T (retires Aug 24)"
-description = "Reasoning model for deliberate analysis, multi-step problem solving, and tool use"
-family = "ring"
-release_date = "2026-05-08"
-last_updated = "2026-05-08"
-attachment = false
-reasoning = true
-temperature = true
-tool_call = true
-structured_output = false
-open_weights = false
-
-[[reasoning_options]]
-type = "effort"
-values = ["high", "xhigh"]
-
-[cost]
-input = 0.3
-output = 2.5
-cache_read = 0.06
-
-[limit]
-context = 262_144
-output = 65_536
-
-[modalities]
-input = ["text"]
-output = ["text"]
@@ -10,3 +10,6 @@ values = ["high"]
 input = 0.3
 output = 1.2
 cache_read = 0.06
+
+[limit]
+context = 196_608
@@ -11,4 +11,4 @@ input = 0.13
 output = 0.52

 [limit]
-output = 8_192
+context = 40_960
@@ -9,5 +9,8 @@ values = ["none", "high"]
 input = 0.26
 output = 2.08

+[limit]
+output = 262_144
+
 [modalities]
 input = ["text", "image", "video"]
@@ -9,5 +9,8 @@ values = ["none", "high"]
 input = 0.39
 output = 2.34

+[limit]
+output = 262_144
+
 [modalities]
 input = ["text", "image", "video"]
@@ -10,7 +10,7 @@ input = 0.45
 output = 2.7

 [limit]
-output = 262_144
+output = 81_920

 [modalities]
 input = ["text", "image", "video"]
@@ -6,10 +6,10 @@ type = "effort"
 values = ["none", "low", "medium", "xhigh"]

 [cost]
-input = 0.5
-output = 3
-cache_read = 0.1
-cache_write = 0.625
+input = 0.425
+output = 2.55
+cache_read = 0.085
+cache_write = 0.53125

 [limit]
 output = 131_072
@@ -0,0 +1,23 @@
+name = "Tencent: Hy-MT2-7B"
+description = "Hy-MT2-7B is a 7B-parameter translation model from Tencent. It supports 33 language pairs and five Chinese dialect and minority-language pairs, with workflows for structured, delimiter-based, contextual, glossary-based, and style-guided translation."
+family = "Hy"
+release_date = "2026-08-19"
+last_updated = "2026-08-19"
+attachment = false
+reasoning = false
+temperature = true
+tool_call = false
+structured_output = true
+open_weights = false
+
+[cost]
+input = 0.074
+output = 0.295
+
+[limit]
+context = 8_192
+output = 4_096
+
+[modalities]
+input = ["text"]
+output = ["text"]
@@ -7,9 +7,9 @@ type = "effort"
 values = ["none", "low", "high"]

 [cost]
-input = 0.132
-output = 0.528
-cache_read = 0.033
+input = 0.14
+output = 0.58
+cache_read = 0.035

 [limit]
 context = 262_144
@@ -14,8 +14,8 @@ input = 0.4
 output = 0.4

 [limit]
-context = 1_024_000
-output = 1_024_000
+context = 32_768
+output = 32_768

 [modalities]
 input = ["text"]
@@ -1,6 +1,6 @@
 base_model = "thinkingmachines/inkling-small"
 description = "Inkling Small is an open-weight multimodal mixture-of-experts model from Thinking Machines Lab, with 12B active parameters out of 276B total. It is positioned as the smaller, more efficient member of..."
-structured_output = true
+structured_output = false

 [[reasoning_options]]
 type = "effort"
@@ -15,7 +15,7 @@ output = 0.65

 [limit]
 context = 6_144
-output = 6_144
+output = 4_096

 [modalities]
 input = ["text"]
@@ -15,9 +15,9 @@ type = "effort"
 values = ["none", "low", "high", "max"]

 [cost]
-input = 0.065
-output = 0.18
-cache_read = 0.02
+input = 0.04
+output = 0.08
+cache_read = 0.008

 [limit]
 context = 1_048_576
@@ -0,0 +1,25 @@
+base_model = "xiaomi/mimo-v2.5-pro"
+name = "MiMo V2.5 Pro (DeepInfra)"
+tool_call = false
+structured_output = false
+
+[interleaved]
+field = "reasoning_content"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "minimal", "low", "medium", "high", "xhigh", "max"]
+
+[cost]
+input = 1
+output = 3
+cache_read = 0.2
+
+[[cost.tiers]]
+tier = { type = "context", size = 256_000 }
+input = 2
+output = 6
+cache_read = 0.4
+
+[limit]
+output = 16_384
@@ -0,0 +1,25 @@
+base_model = "xiaomi/mimo-v2.5"
+name = "MiMo V2.5 (DeepInfra)"
+structured_output = true
+
+[interleaved]
+field = "reasoning_content"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "minimal", "low", "medium", "high", "xhigh", "max"]
+
+[cost]
+input = 0.4
+output = 2
+cache_read = 0.08
+
+[[cost.tiers]]
+tier = { type = "context", size = 256_000 }
+input = 0.8
+output = 4
+cache_read = 0.16
+
+[limit]
+context = 262_144
+output = 16_384
@@ -0,0 +1,21 @@
+base_model = "xiaomi/mimo-v2.5-pro"
+name = "MiMo V2.5 Pro (NovitaAI)"
+structured_output = false
+
+[interleaved]
+field = "reasoning_content"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "minimal", "low", "medium", "high"]
+
+[cost]
+input = 0.522
+output = 1.044
+cache_read = 0.0043
+
+[[cost.tiers]]
+tier = { type = "context", size = 256_000 }
+input = 2
+output = 6
+cache_read = 0.4
@@ -0,0 +1,21 @@
+base_model = "xiaomi/mimo-v2.5"
+name = "MiMo V2.5 (NovitaAI)"
+structured_output = false
+
+[interleaved]
+field = "reasoning_content"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "low", "medium", "high"]
+
+[cost]
+input = 0.168
+output = 0.336
+cache_read = 0.0034
+
+[[cost.tiers]]
+tier = { type = "context", size = 256_000 }
+input = 0.8
+output = 4
+cache_read = 0.16
@@ -1,23 +0,0 @@
-name = "GPT-4o Mini Search Preview (OpenAI)"
-description = "Compact GPT model for low-latency assistance and high-volume workloads"
-family = "gpt"
-release_date = "2024-10-01"
-last_updated = "2024-10-01"
-attachment = true
-reasoning = false
-temperature = false
-tool_call = false
-structured_output = false
-open_weights = false
-
-[cost]
-input = 0.15
-output = 0.6
-
-[limit]
-context = 128_000
-output = 16_384
-
-[modalities]
-input = ["text", "image"]
-output = ["text"]
@@ -1,23 +0,0 @@
-name = "GPT-4o Search Preview (OpenAI)"
-description = "GPT model for general reasoning, writing, coding, and tool-assisted tasks"
-family = "gpt"
-release_date = "2024-10-01"
-last_updated = "2024-10-01"
-attachment = true
-reasoning = false
-temperature = false
-tool_call = false
-structured_output = false
-open_weights = false
-
-[cost]
-input = 2.5
-output = 10
-
-[limit]
-context = 128_000
-output = 16_384
-
-[modalities]
-input = ["text", "image"]
-output = ["text"]
@@ -1,23 +0,0 @@
-name = "GPT-4o Mini Search Preview"
-description = "Compact GPT model for low-latency assistance and high-volume workloads"
-family = "gpt"
-release_date = "2024-10-01"
-last_updated = "2024-10-01"
-attachment = true
-reasoning = false
-temperature = true
-tool_call = false
-structured_output = false
-open_weights = false
-
-[cost]
-input = 0.15
-output = 0.60
-
-[limit]
-context = 128_000
-output = 16_384
-
-[modalities]
-input = ["text", "image"]
-output = ["text"]
@@ -1,23 +0,0 @@
-name = "GPT-4o Search Preview"
-description = "GPT model for general reasoning, writing, coding, and tool-assisted tasks"
-family = "gpt"
-release_date = "2024-10-01"
-last_updated = "2024-10-01"
-attachment = true
-reasoning = false
-temperature = true
-tool_call = false
-structured_output = false
-open_weights = false
-
-[cost]
-input = 2.50
-output = 10.00
-
-[limit]
-context = 128_000
-output = 16_384
-
-[modalities]
-input = ["text", "image"]
-output = ["text"]
@@ -8,7 +8,7 @@ field = "reasoning_content"

 [[reasoning_options]]
 type = "effort"
-values = ["none", "low", "medium", "high"]
+values = ["none", "minimal", "low", "medium", "high", "xhigh", "max"]

 [cost]
 input = 0.435
@@ -20,6 +20,3 @@ tier = { type = "context", size = 256_000 }
 input = 2
 output = 6
 cache_read = 0.4
-
-[limit]
-context = 1_000_000
@@ -8,7 +8,7 @@ field = "reasoning_content"

 [[reasoning_options]]
 type = "effort"
-values = ["none", "low", "medium", "high"]
+values = ["none", "minimal", "low", "medium", "high", "xhigh", "max"]

 [cost]
 input = 0.14
@@ -20,6 +20,3 @@ tier = { type = "context", size = 256_000 }
 input = 0.8
 output = 4
 cache_read = 0.16
-
-[limit]
-context = 1_000_000
@@ -9,6 +9,6 @@ type = "effort"
 values = ["none", "low", "medium", "high", "max"]

 [cost]
-input = 0.22
-output = 0.66
+input = 0.035
+output = 0.07
 cache_read = 0.007
@@ -0,0 +1,25 @@
+name = "Gemma 4 31B MeroMero v2"
+description = "Gemma 4 31B MeroMero v2 is a LoRA finetune for emotive dialogue, relationship scenes, creative writing, and multimodal roleplay."
+family = "gemma"
+release_date = "2026-08-23"
+last_updated = "2026-08-23"
+attachment = true
+reasoning = true
+tool_call = true
+structured_output = false
+open_weights = true
+reasoning_options = []
+
+[cost]
+input = 0.08
+output = 0.33
+cache_read = 0.04
+
+[limit]
+context = 65_536
+input = 65_536
+output = 32_768
+
+[modalities]
+input = ["text", "image"]
+output = ["text"]
@@ -1,24 +0,0 @@
-name = "Amazon Nova Micro 1.0"
-description = "Efficient model for low-latency assistance, extraction, and routine automation"
-family = "nova-micro"
-release_date = "2024-01-01"
-last_updated = "2024-12-03"
-attachment = false
-reasoning = false
-tool_call = false
-structured_output = false
-open_weights = false
-
-[cost]
-input = 0.0357
-output = 0.1394
-cache_read = 0.01785
-
-[limit]
-context = 128_000
-input = 128_000
-output = 5_120
-
-[modalities]
-input = ["text"]
-output = ["text"]
@@ -1,22 +1,16 @@
 # Not included in subscription
 # Tracks Gemini 3.5 Flash (current flash-latest target).
 base_model = "google/gemini-flash-latest"
-tool_call = false
-structured_output = false

 [[reasoning_options]]
 type = "effort"
 values = ["minimal", "low", "medium", "high"]

 [cost]
-input = 0.75
-output = 3.75
-cache_read = 0.075
-cache_write = 0.041667
+input = 0.375
+output = 1.875
+cache_read = 0.0375
+cache_write = 0.020833

 [limit]
-context = 1_048_756
-input = 1_048_756
-
-[modalities]
-input = ["text", "image", "audio"]
+input = 1_048_576
@@ -0,0 +1,24 @@
+name = "Gemma 4 26B A4B Uncensored"
+description = "Gemma 4 26B A4B Uncensored is an NVFP4 open-weight multimodal mixture-of-experts model tuned for fewer refusals across chat, coding, tool use, and long-context work."
+family = "gemma"
+release_date = "2026-08-22"
+last_updated = "2026-08-22"
+attachment = true
+reasoning = false
+tool_call = true
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.08
+output = 0.33
+cache_read = 0.04
+
+[limit]
+context = 131_072
+input = 131_072
+output = 32_768
+
+[modalities]
+input = ["text", "image"]
+output = ["text"]
@@ -0,0 +1,25 @@
+name = "Ornith 1.5 9B"
+description = "Ornith 1.5 9B is an FP8 dense open-weight reasoning model built for agentic coding, tool use, visual understanding, and efficient long-context work."
+family = "ornith"
+release_date = "2026-08-23"
+last_updated = "2026-08-23"
+attachment = true
+reasoning = true
+tool_call = true
+structured_output = true
+open_weights = true
+reasoning_options = []
+
+[cost]
+input = 0.05
+output = 0.1
+cache_read = 0.025
+
+[limit]
+context = 131_072
+input = 131_072
+output = 32_768
+
+[modalities]
+input = ["text", "image"]
+output = ["text"]
@@ -0,0 +1,17 @@
+# NIM Chat schema: `reasoning_effort = none|high|max`; `none` disables thinking.
+# https://docs.api.nvidia.com/nim/reference/deepseek-ai-deepseek-v4-flash-infer
+#
+# Pricing: hosted on NVIDIA's API trial tier and currently free — no separate
+# list rate is published for this ID. Source: this model's catalog card,
+# https://build.nvidia.com/deepseek-ai/deepseek-v4-flash-0731 (governed by the
+# NVIDIA API Trial Terms of Service).
+base_model = "deepseek/deepseek-v4-flash-0731"
+
+reasoning_options = [{ type = "effort", values = ["none", "high", "max"] }]
+
+[interleaved]
+field = "reasoning_content"
+
+[cost]
+input = 0.0
+output = 0.0
@@ -0,0 +1,16 @@
+# NIM wire syntax (verified against integrate.api.nvidia.com/v1, 2026-08-22):
+#   on/off: chat_template_kwargs = { "thinking": true | false }  (default: true)
+#   effort: top-level "reasoning_effort" = "low" | "high" | "max"
+base_model = "moonshotai/kimi-k3"
+
+reasoning_options = [
+  { type = "toggle" },
+  { type = "effort", values = ["low", "high", "max"] },
+]
+
+[interleaved]
+field = "reasoning_content"
+
+[cost]
+input = 0.0
+output = 0.0
@@ -0,0 +1,17 @@
+# https://ofox.ai/models/deepseek/deepseek-v4-flash-0731
+# https://api-docs.deepseek.com/quick_start/pricing/
+# Toggle: thinking.type = enabled|disabled
+# Effort: reasoning_effort = low|high|max (Flash; lab maps xhigh→high)
+base_model = "deepseek/deepseek-v4-flash-0731"
+
+[[reasoning_options]]
+type = "toggle"
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "high", "max"]
+
+[cost]
+input = 0.44
+output = 1.32
+cache_read = 0.014
@@ -0,0 +1,10 @@
+base_model = "meituan/longcat-2.0"
+reasoning_options = [{ type = "toggle" }]
+
+[interleaved]
+field = "reasoning_content"
+
+[cost]
+input = 0.30
+output = 1.20
+cache_read = 0.006
@@ -8,9 +8,10 @@ type = "effort"
 values = ["low", "high", "max"]

 [cost]
-input = 0.08
-output = 0.18
-cache_read = 0.016
+input = 0.14
+output = 0.28
+cache_read = 0.028

 [limit]
 context = 1_310_720
+output = 131_072
@@ -11,9 +11,9 @@ type = "effort"
 values = ["high", "xhigh"]

 [cost]
-input = 0.06006
-output = 0.12012
-cache_read = 0.012012
+input = 0.0574
+output = 0.1148
+cache_read = 0.01148

 [limit]
 context = 1_048_576
@@ -8,9 +8,9 @@ type = "effort"
 values = ["low", "high", "max"]

 [cost]
-input = 1.188
-output = 3.564
-cache_read = 0.0396
+input = 1.122
+output = 3.366
+cache_read = 0.0374

 [limit]
 context = 1_048_576
@@ -11,9 +11,9 @@ type = "effort"
 values = ["high", "xhigh"]

 [cost]
-input = 0.413772
-output = 0.827544
-cache_read = 0.034481
+input = 0.526176
+output = 1.052352
+cache_read = 0.043848

 [limit]
 context = 1_048_576
@@ -1,24 +0,0 @@
-name = "Ling-2.6-1T"
-description = "Tool-capable chat model for instruction following and agentic application workflows"
-family = "ling"
-release_date = "2026-04-23"
-last_updated = "2026-04-23"
-attachment = false
-reasoning = false
-temperature = true
-tool_call = true
-structured_output = true
-open_weights = false
-
-[cost]
-input = 0.075
-output = 0.625
-cache_read = 0.015
-
-[limit]
-context = 262_144
-output = 32_768
-
-[modalities]
-input = ["text"]
-output = ["text"]
@@ -1,24 +0,0 @@
-name = "Ling-2.6-flash"
-description = "Efficient model for low-latency assistance, extraction, and routine automation"
-family = "ling"
-release_date = "2026-04-21"
-last_updated = "2026-04-21"
-attachment = false
-reasoning = false
-temperature = true
-tool_call = true
-structured_output = true
-open_weights = false
-
-[cost]
-input = 0.01
-output = 0.03
-cache_read = 0.002
-
-[limit]
-context = 262_144
-output = 32_768
-
-[modalities]
-input = ["text"]
-output = ["text"]
@@ -1,28 +0,0 @@
-name = "Ring-2.6-1T"
-description = "Reasoning model for deliberate analysis, multi-step problem solving, and tool use"
-family = "ring"
-release_date = "2026-05-08"
-last_updated = "2026-05-08"
-attachment = false
-reasoning = true
-temperature = true
-tool_call = true
-structured_output = false
-open_weights = false
-
-[[reasoning_options]]
-type = "effort"
-values = ["high", "xhigh"]
-
-[cost]
-input = 0.075
-output = 0.625
-cache_read = 0.015
-
-[limit]
-context = 262_144
-output = 65_536
-
-[modalities]
-input = ["text"]
-output = ["text"]
@@ -10,6 +10,6 @@ type = "effort"
 values = ["low", "medium", "high", "xhigh"]

 [cost]
-input = 0.3
-output = 1.1
+input = 0.35
+output = 1.5
 cache_read = 0.04
@@ -3,6 +3,6 @@ structured_output = true
 reasoning_options = []

 [cost]
-input = 0.3
-output = 1.2
-cache_read = 0.06
+input = 0.24
+output = 0.96
+cache_read = 0.048
@@ -5,9 +5,9 @@ reasoning_options = []
 field = "reasoning_details"

 [cost]
-input = 0.5415
-output = 2.28
-cache_read = 0.0912
+input = 0.95
+output = 4
+cache_read = 0.16

 [modalities]
 input = ["text", "image"]
@@ -5,9 +5,8 @@ type = "effort"
 values = ["low", "medium", "high"]

 [cost]
-input = 0.03
+input = 0.037
 output = 0.17
-cache_read = 0.03

 [limit]
 output = 131_072
@@ -6,8 +6,5 @@ structured_output = false
 type = "toggle"

 [cost]
-input = 0.13
-output = 0.52
-
-[limit]
-output = 8_192
+input = 0.12
+output = 0.5
@@ -7,5 +7,8 @@ type = "toggle"
 input = 0.26
 output = 2.08

+[limit]
+output = 262_144
+
 [modalities]
 input = ["text", "image", "video"]
@@ -4,8 +4,12 @@ base_model = "alibaba/qwen3.5-397b-a17b"
 type = "toggle"

 [cost]
-input = 0.39
-output = 2.34
+input = 0.5
+output = 3.6
+cache_read = 0.3
+
+[limit]
+output = 262_144

 [modalities]
 input = ["text", "image", "video"]
@@ -4,12 +4,11 @@ base_model = "alibaba/qwen3.6-27b"
 type = "toggle"

 [cost]
-input = 0.6
-output = 3.6
-cache_read = 0.12
+input = 0.32
+output = 3.2

 [limit]
-output = 262_144
+output = 81_920

 [modalities]
 input = ["text", "image", "video"]
@@ -9,8 +9,8 @@ type = "effort"
 values = ["low", "medium", "xhigh"]

 [cost]
-input = 0.45
-output = 3.2
+input = 0.4
+output = 3
 cache_read = 0.05

 [limit]
@@ -0,0 +1,23 @@
+name = "Hy-MT2-7B"
+description = "Tencent Hy reasoning model for coding, instruction following, and agent tasks"
+family = "Hy"
+release_date = "2026-08-19"
+last_updated = "2026-08-19"
+attachment = false
+reasoning = false
+temperature = true
+tool_call = false
+structured_output = true
+open_weights = true
+
+[cost]
+input = 0.074
+output = 0.295
+
+[limit]
+context = 8_192
+output = 4_096
+
+[modalities]
+input = ["text"]
+output = ["text"]
@@ -16,7 +16,7 @@ output = 0.4

 [limit]
 context = 1_024_000
-output = 1_024_000
+output = 32_768

 [modalities]
 input = ["text"]
@@ -1,6 +1,6 @@
 base_model = "thinkingmachines/inkling-small"
 description = "Multimodal reasoning model for visual analysis, planning, and tool use"
-structured_output = true
+structured_output = false

 [[reasoning_options]]
 type = "effort"
@@ -7,9 +7,9 @@ type = "effort"
 values = ["none", "minimal", "low", "medium", "high", "max"]

 [cost]
-input = 0.95
+input = 1
 output = 4.05
-cache_read = 0.16
+cache_read = 0.17

 [limit]
 output = 262_144
@@ -16,7 +16,7 @@ output = 0.65

 [limit]
 context = 6_144
-output = 6_144
+output = 4_096

 [modalities]
 input = ["text"]
@@ -18,9 +18,9 @@ type = "effort"
 values = ["low", "high", "max"]

 [cost]
-input = 0.065
-output = 0.18
-cache_read = 0.02
+input = 0.04
+output = 0.08
+cache_read = 0.008

 [limit]
 context = 1_310_720
@@ -0,0 +1,3 @@
+<svg xmlns="http://www.w3.org/2000/svg" viewBox="0 0 315 315" fill="currentColor">
+  <path fill-rule="evenodd" clip-rule="evenodd" d="M159.78 315C71.53 315 0 244.49 0 157.5C0 -18.9499 159.78 0.650075 159.78 0.650075C159.78 87.2201 88.36 157.4 0.2 157.5C149.8 157.64 159.78 315 159.78 315ZM160.52 217.98C160.52 217.98 156.94 161.65 105.04 157.52C120.6 157.34 160.52 151.54 160.52 96.5601C160.52 151.54 200.44 157.34 216 157.52C164.1 161.63 160.52 217.98 160.52 217.98Z"/>
+</svg>
@@ -0,0 +1,15 @@
+base_model = "anthropic/claude-fable-5"
+structured_output = true
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high", "xhigh", "max"]
+
+[cost]
+input = 10
+output = 50
+cache_read = 1
+cache_write = 12.5
+
+[interleaved]
+field = "reasoning_content"
@@ -0,0 +1,13 @@
+base_model = "anthropic/claude-haiku-4-5"
+structured_output = true
+
+reasoning_options = []
+
+[cost]
+input = 1
+output = 5
+cache_read = 0.1
+cache_write = 1.25
+
+[interleaved]
+field = "reasoning_content"
@@ -0,0 +1,15 @@
+base_model = "anthropic/claude-opus-4-5"
+structured_output = true
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high"]
+
+[cost]
+input = 5
+output = 25
+cache_read = 0.5
+cache_write = 6.25
+
+[interleaved]
+field = "reasoning_content"
@@ -0,0 +1,15 @@
+base_model = "anthropic/claude-opus-4-6"
+structured_output = true
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high", "max"]
+
+[cost]
+input = 5
+output = 25
+cache_read = 0.5
+cache_write = 6.25
+
+[interleaved]
+field = "reasoning_content"
@@ -0,0 +1,15 @@
+base_model = "anthropic/claude-opus-4-7"
+structured_output = true
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high", "xhigh", "max"]
+
+[cost]
+input = 5
+output = 25
+cache_read = 0.5
+cache_write = 6.25
+
+[interleaved]
+field = "reasoning_content"
@@ -0,0 +1,15 @@
+base_model = "anthropic/claude-opus-4-8"
+structured_output = true
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high", "xhigh", "max"]
+
+[cost]
+input = 5
+output = 25
+cache_read = 0.5
+cache_write = 6.25
+
+[interleaved]
+field = "reasoning_content"
@@ -0,0 +1,15 @@
+base_model = "anthropic/claude-opus-5"
+structured_output = true
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high", "xhigh", "max"]
+
+[cost]
+input = 5
+output = 25
+cache_read = 0.5
+cache_write = 6.25
+
+[interleaved]
+field = "reasoning_content"
@@ -0,0 +1,13 @@
+base_model = "anthropic/claude-sonnet-4-5"
+structured_output = true
+
+reasoning_options = []
+
+[cost]
+input = 3
+output = 15
+cache_read = 0.3
+cache_write = 3.75
+
+[interleaved]
+field = "reasoning_content"
@@ -0,0 +1,15 @@
+base_model = "anthropic/claude-sonnet-4-6"
+structured_output = true
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high", "max"]
+
+[cost]
+input = 3
+output = 15
+cache_read = 0.3
+cache_write = 3.75
+
+[interleaved]
+field = "reasoning_content"
@@ -0,0 +1,15 @@
+base_model = "anthropic/claude-sonnet-5"
+structured_output = true
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high", "xhigh", "max"]
+
+[cost]
+input = 2
+output = 10
+cache_read = 0.2
+cache_write = 2.5
+
+[interleaved]
+field = "reasoning_content"
@@ -0,0 +1,10 @@
+base_model = "google/gemini-3-flash-preview"
+
+[[reasoning_options]]
+type = "effort"
+values = ["minimal", "low", "medium", "high"]
+
+[cost]
+input = 0.5
+output = 3
+cache_read = 0.05
--- a/Show More
+++ b/Show More