fix: update GPT-5.6 Terra/Luna pricing for OpenAI and Azure (#3884)

OpenAI cut Terra 20% and Luna 80% on 2026-07-30. Update standard,
long-context tier, and fast-mode costs for openai, azure, and
azure-cognitive-services. Sol unchanged.

providers/azure-cognitive-services/models/gpt-5.6-luna.toml

@@ -1,14 +1,15 @@
+# Pricing aligned with OpenAI standard (Terra/Luna cut 2026-07-30): https://developers.openai.com/api/docs/pricing
 base_model = "openai/gpt-5.6-luna"
 status = "beta"
 reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }]
 
 [cost]
-input = 1
-output = 6
-cache_read = 0.1
+input = 0.2
+output = 1.2
+cache_read = 0.02
 
 [[cost.tiers]]
 tier = { type = "context", size = 272_000 }
-input = 2
-output = 9
-cache_read = 0.2
+input = 0.4
+output = 1.8
+cache_read = 0.04

providers/azure-cognitive-services/models/gpt-5.6-terra.toml

@@ -1,14 +1,15 @@
+# Pricing aligned with OpenAI standard (Terra/Luna cut 2026-07-30): https://developers.openai.com/api/docs/pricing
 base_model = "openai/gpt-5.6-terra"
 status = "beta"
 reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }]
 
 [cost]
-input = 2.5
-output = 15
-cache_read = 0.25
+input = 2.0
+output = 12
+cache_read = 0.2
 
 [[cost.tiers]]
 tier = { type = "context", size = 272_000 }
-input = 5
-output = 22.5
-cache_read = 0.5
+input = 4
+output = 18
+cache_read = 0.4

providers/azure/models/gpt-5.6-luna.toml

@@ -1,14 +1,15 @@
+# Pricing aligned with OpenAI standard (Terra/Luna cut 2026-07-30): https://developers.openai.com/api/docs/pricing
 base_model = "openai/gpt-5.6-luna"
 status = "beta"
 reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }]
 
 [cost]
-input = 1
-output = 6
-cache_read = 0.1
+input = 0.2
+output = 1.2
+cache_read = 0.02
 
 [[cost.tiers]]
 tier = { type = "context", size = 272_000 }
-input = 2
-output = 9
-cache_read = 0.2
+input = 0.4
+output = 1.8
+cache_read = 0.04

providers/azure/models/gpt-5.6-terra.toml

@@ -1,14 +1,15 @@
+# Pricing aligned with OpenAI standard (Terra/Luna cut 2026-07-30): https://developers.openai.com/api/docs/pricing
 base_model = "openai/gpt-5.6-terra"
 status = "beta"
 reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }]
 
 [cost]
-input = 2.5
-output = 15
-cache_read = 0.25
+input = 2.0
+output = 12
+cache_read = 0.2
 
 [[cost.tiers]]
 tier = { type = "context", size = 272_000 }
-input = 5
-output = 22.5
-cache_read = 0.5
+input = 4
+output = 18
+cache_read = 0.4

providers/openai/models/gpt-5.6-luna.toml

@@ -1,21 +1,22 @@
+# Pricing: https://developers.openai.com/api/docs/pricing (Terra/Luna cut 2026-07-30)
 base_model = "openai/gpt-5.6-luna"
 reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }]
 
 [cost]
-input = 1.00
-output = 6.00
-cache_read = 0.10
-cache_write = 1.25
+input = 0.20
+output = 1.20
+cache_read = 0.02
+cache_write = 0.25
 
 [[cost.tiers]]
 tier = { size = 272_000 }
-input = 2.00
-output = 9.00
-cache_read = 0.20
-cache_write = 2.50
+input = 0.40
+output = 1.80
+cache_read = 0.04
+cache_write = 0.50
 
 [experimental.modes.fast]
-cost = { input = 2.00, output = 12.00, cache_read = 0.20, cache_write = 2.50 }
+cost = { input = 0.40, output = 2.40, cache_read = 0.04, cache_write = 0.50 }
 provider = { body = { service_tier = "priority" } }
 
 [experimental.modes.pro]

providers/openai/models/gpt-5.6-terra.toml

@@ -1,21 +1,22 @@
+# Pricing: https://developers.openai.com/api/docs/pricing (Terra/Luna cut 2026-07-30)
 base_model = "openai/gpt-5.6-terra"
 reasoning_options = [{ type = "effort", values = ["none", "low", "medium", "high", "xhigh", "max"] }]
 
 [cost]
-input = 2.50
-output = 15.00
-cache_read = 0.25
-cache_write = 3.125
+input = 2.00
+output = 12.00
+cache_read = 0.20
+cache_write = 2.50
 
 [[cost.tiers]]
 tier = { size = 272_000 }
-input = 5.00
-output = 22.50
-cache_read = 0.50
-cache_write = 6.25
+input = 4.00
+output = 18.00
+cache_read = 0.40
+cache_write = 5.00
 
 [experimental.modes.fast]
-cost = { input = 5.00, output = 30.00, cache_read = 0.50, cache_write = 6.25 }
+cost = { input = 4.00, output = 24.00, cache_read = 0.40, cache_write = 5.00 }
 provider = { body = { service_tier = "priority" } }
 
 [experimental.modes.pro]