feat(aixy): add curated 50-model catalog

providers/aixy/models/alibaba/glm-5.2.toml

@@ -0,0 +1,20 @@
+# Sources (accessed 2026-08-08):
+# https://www.alibabacloud.com/help/en/model-studio/glm
+# https://www.alibabacloud.com/help/en/model-studio/model-pricing (Singapore list)
+# https://www.qwencloud.com/models/glm-5.2 (implicit cache price)
+# Effort: reasoning_effort = none|minimal|low|medium|high|xhigh|max
+# (none disables reasoning, so a separate toggle is not needed).
+base_model = "zhipuai/glm-5.2"
+
+[interleaved]
+field = "reasoning_content"
+
+[[reasoning_options]]
+type = "effort"
+values = ["none", "minimal", "low", "medium", "high", "xhigh", "max"]
+
+[cost]
+input = 1.4
+output = 4.4
+cache_read = 0.28
+cache_write = 0

providers/aixy/models/alibaba/qwen3.6-27b.toml

@@ -0,0 +1,14 @@
+# Toggle: enable_thinking = true|false
+# Aixy forwards this route to alibaba without translating the model payload.
+# Reasoning controls mirror providers/alibaba/models/qwen3.6-27b.toml.
+base_model = "alibaba/qwen3.6-27b"
+
+[[reasoning_options]]
+type = "toggle"
+
+[[reasoning_options]]
+type = "budget_tokens"
+
+[cost]
+input = 0.6
+output = 3.6

providers/aixy/models/alibaba/qwen3.6-flash.toml

@@ -0,0 +1,15 @@
+# Toggle: enable_thinking = true|false
+# Aixy forwards this route to alibaba without translating the model payload.
+# Reasoning controls mirror providers/alibaba/models/qwen3.6-flash.toml.
+base_model = "alibaba/qwen3.6-flash"
+
+[[reasoning_options]]
+type = "toggle"
+
+[[reasoning_options]]
+type = "budget_tokens"
+
+[cost]
+input = 0.1875
+output = 1.125
+cache_write = 0.234375

providers/aixy/models/alibaba/qwen3.7-max.toml

@@ -0,0 +1,16 @@
+# Toggle: enable_thinking = true|false
+# Aixy forwards this route to alibaba without translating the model payload.
+# Reasoning controls mirror providers/alibaba/models/qwen3.7-max.toml.
+base_model = "alibaba/qwen3.7-max"
+
+[[reasoning_options]]
+type = "toggle"
+
+[[reasoning_options]]
+type = "budget_tokens"
+
+[cost]
+input = 2.5
+output = 7.5
+cache_read = 0.5
+cache_write = 3.125

providers/aixy/models/alibaba/qwen3.7-plus.toml

@@ -0,0 +1,27 @@
+# Toggle: enable_thinking = true|false
+# Aixy forwards this route to alibaba without translating the model payload.
+# Reasoning controls mirror providers/alibaba/models/qwen3.7-plus.toml.
+base_model = "alibaba/qwen3.7-plus"
+attachment = false
+
+[[reasoning_options]]
+type = "toggle"
+
+[[reasoning_options]]
+type = "budget_tokens"
+
+[cost]
+input = 0.5
+output = 3
+cache_read = 0.05
+cache_write = 0.625
+
+[[cost.tiers]]
+tier = { type = "context", size = 256_000 }
+input = 2
+output = 6
+cache_read = 0.2
+cache_write = 2.5
+
+[limit]
+output = 65_536

providers/aixy/models/alibaba/qwen3.8-max.toml

@@ -0,0 +1,34 @@
+# Sources (accessed 2026-08-04):
+# https://www.qwencloud.com/models/qwen3.8-max
+# https://docs.qwencloud.com/developer-guides/text-generation/thinking
+# https://docs.qwencloud.com/developer-guides/getting-started/text-generation-models
+# https://help.aliyun.com/zh/model-studio/model-pricing (Singapore: qwen3.8-max list)
+# Toggle: enable_thinking true|false (hybrid)
+# Effort: reasoning_effort = low|medium|xhigh (default xhigh); high accepted as alias → xhigh
+# Budget: thinking_budget (0..262144) cannot be combined with reasoning_effort
+# API: {"enable_thinking":true,"reasoning_effort":"medium"} or
+# {"enable_thinking":true,"thinking_budget":16384}
+# Pay-as-you-go on DashScope/QwenCloud (not Token Plan only). Cost USD/MTok from model page.
+base_model = "alibaba/qwen3.8-max"
+structured_output = true
+
+[interleaved]
+field = "reasoning_content"
+
+[[reasoning_options]]
+type = "toggle"
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "xhigh"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+min = 0
+max = 262_144
+
+[cost]
+input = 2
+output = 6
+cache_read = 0.25
+cache_write = 2.5

providers/aixy/models/anthropic/claude-fable-5.toml

@@ -0,0 +1,12 @@
+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

providers/aixy/models/anthropic/claude-haiku-4-5.toml

@@ -0,0 +1,12 @@
+base_model = "anthropic/claude-haiku-4-5"
+structured_output = true
+
+[[reasoning_options]]
+type = "budget_tokens"
+min = 1_024
+
+[cost]
+input = 1
+output = 5
+cache_read = 0.1
+cache_write = 1.25

providers/aixy/models/anthropic/claude-opus-4-6.toml

@@ -0,0 +1,16 @@
+base_model = "anthropic/claude-opus-4-6"
+structured_output = true
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+min = 1_024
+
+[cost]
+input = 5
+output = 25
+cache_read = 0.5
+cache_write = 6.25

providers/aixy/models/anthropic/claude-opus-4-7.toml

@@ -0,0 +1,12 @@
+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

providers/aixy/models/anthropic/claude-opus-4-8.toml

@@ -0,0 +1,16 @@
+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
+
+[experimental.modes.fast]
+cost = { input = 10, output = 50, cache_read = 1, cache_write = 12.5 }
+provider = { body = { speed = "fast" }, headers = { anthropic-beta = "fast-mode-2026-02-01" } }

providers/aixy/models/anthropic/claude-opus-5.toml

@@ -0,0 +1,16 @@
+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
+
+[experimental.modes.fast]
+cost = { input = 10, output = 50, cache_read = 1, cache_write = 12.5 }
+provider = { body = { speed = "fast" }, headers = { anthropic-beta = "fast-mode-2026-02-01" } }

providers/aixy/models/anthropic/claude-sonnet-4-6.toml

@@ -0,0 +1,19 @@
+base_model = "anthropic/claude-sonnet-4-6"
+structured_output = true
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high", "max"]
+
+[[reasoning_options]]
+type = "budget_tokens"
+min = 1_024
+
+[cost]
+input = 3
+output = 15
+cache_read = 0.3
+cache_write = 3.75
+
+[limit]
+output = 128_000

providers/aixy/models/anthropic/claude-sonnet-5.toml

@@ -0,0 +1,18 @@
+# Toggle: thinking.type = adaptive|disabled
+# Aixy forwards this route to anthropic without translating the model payload.
+# Reasoning controls mirror providers/anthropic/models/claude-sonnet-5.toml.
+base_model = "anthropic/claude-sonnet-5"
+structured_output = true
+
+[[reasoning_options]]
+type = "toggle"
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high", "xhigh", "max"]
+
+[cost]
+input = 2
+output = 10
+cache_read = 0.2
+cache_write = 2.5

providers/aixy/models/deepseek/deepseek-v4-flash.toml

@@ -0,0 +1,26 @@
+# Toggle: thinking.type = enabled|disabled
+# Reasoning tokens are billed at the output rate (no separate CoT price).
+# `completion_tokens_details.reasoning_tokens` is a subset of completion_tokens.
+# https://api-docs.deepseek.com/quick_start/pricing/ (accessed 2026-07-31)
+# OpenAI: `thinking.type = enabled|disabled`, `reasoning_effort = low|high|max`.
+# Anthropic: `thinking.type`, `output_config.effort = low|high|max`; budget ignored.
+# Flash maps requested low→low (unlike Pro, which maps low→high). xhigh→high.
+# https://api-docs.deepseek.com/guides/thinking_mode/ (accessed 2026-08-02)
+base_model = "deepseek/deepseek-v4-flash-0731"
+name = "DeepSeek V4 Flash"
+
+[interleaved]
+field = "reasoning_content"
+
+[[reasoning_options]]
+type = "toggle"
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "high", "max"]
+
+[cost]
+input = 0.14
+output = 0.28
+reasoning = 0.28
+cache_read = 0.0028

providers/aixy/models/deepseek/deepseek-v4-pro.toml

@@ -0,0 +1,22 @@
+# Toggle: thinking.type = enabled|disabled
+# Reasoning tokens are billed at the output rate (no separate CoT price).
+# `completion_tokens_details.reasoning_tokens` is a subset of completion_tokens.
+# https://api-docs.deepseek.com/quick_start/pricing/ (accessed 2026-08-12)
+base_model = "deepseek/deepseek-v4-pro-0813"
+name = "DeepSeek V4 Pro"
+
+[interleaved]
+field = "reasoning_content"
+
+[[reasoning_options]]
+type = "toggle"
+
+[[reasoning_options]]
+type = "effort"
+values = ["high", "max"]
+
+[cost]
+input = 0.435
+output = 0.87
+reasoning = 0.87
+cache_read = 0.003625

providers/aixy/models/gemini/gemini-2.5-pro.toml

@@ -0,0 +1,17 @@
+base_model = "google/gemini-2.5-pro"
+
+[[reasoning_options]]
+type = "budget_tokens"
+min = 128
+max = 32_768
+
+[cost]
+input = 1.25
+output = 10
+cache_read = 0.125
+
+[[cost.tiers]]
+tier = { type = "context", size = 200_000 }
+input = 2.5
+output = 15
+cache_read = 0.25

providers/aixy/models/gemini/gemini-3.1-flash-lite.toml

@@ -0,0 +1,11 @@
+base_model = "google/gemini-3.1-flash-lite"
+
+[[reasoning_options]]
+type = "effort"
+values = ["minimal", "low", "medium", "high"]
+
+[cost]
+input = 0.25
+output = 1.5
+cache_read = 0.025
+input_audio = 0.5

providers/aixy/models/gemini/gemini-3.5-flash-lite.toml

@@ -0,0 +1,13 @@
+# Sources:
+# - https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-3-6-flash-3-5-flash-lite-3-5-flash-cyber/
+# - https://ai.google.dev/gemini-api/docs/pricing
+base_model = "google/gemini-3.5-flash-lite"
+
+[[reasoning_options]]
+type = "effort"
+values = ["minimal", "low", "medium", "high"]
+
+[cost]
+input = 0.3
+output = 2.5
+cache_read = 0.03

providers/aixy/models/gemini/gemini-3.5-flash.toml

@@ -0,0 +1,11 @@
+base_model = "google/gemini-3.5-flash"
+
+[[reasoning_options]]
+type = "effort"
+values = ["minimal", "low", "medium", "high"]
+
+[cost]
+input = 1.5
+output = 9
+cache_read = 0.15
+input_audio = 1.5

providers/aixy/models/gemini/gemini-3.6-flash.toml

@@ -0,0 +1,15 @@
+# Sources:
+# - https://ai.google.dev/gemini-api/docs/pricing#gemini-3.6-flash
+# Introductory Standard pricing applies through 2026-12-31. The published
+# 2027 rates are $1.50 input, $7.50 output, and $0.15 cache read.
+base_model = "google/gemini-3.6-flash"
+
+[[reasoning_options]]
+type = "effort"
+values = ["minimal", "low", "medium", "high"]
+
+[cost]
+input = 0.75
+output = 3.75
+cache_read = 0.075
+input_audio = 0.75

providers/aixy/models/gemini/gemini-3.7-flash.toml

@@ -0,0 +1,15 @@
+# Sources:
+# - https://blog.google/innovation-and-ai/models-and-research/gemini-models/introducing-gemini-3-7-flash/
+# - https://ai.google.dev/gemini-api/docs/pricing
+# - https://docs.cloud.google.com/gemini-enterprise-agent-platform/models/thinking
+base_model = "google/gemini-3.7-flash"
+
+[[reasoning_options]]
+type = "effort"
+values = ["low", "medium", "high"]
+
+[cost]
+input = 0.75
+output = 3.75
+cache_read = 0.075
+input_audio = 0.75