From 1d187b5dc0cad400030a637297644ca0684a3cd8 Mon Sep 17 00:00:00 2001
From: Supernova <supernova@appfolio.com>
Date: Fri, 7 Aug 2026 19:36:08 +0000
Subject: [PATCH] Update GPT-5.6 Luna/Terra Bedrock pricing for July 30, 2026
 price cut

AWS Bedrock and OpenAI cut prices for the GPT-5.6 family effective
July 30, 2026. Update the two affected Bedrock entries:

  Luna:  input 1.00 -> 0.22, output 6.00 -> 1.32 (per million tokens)
  Terra: input 2.50 -> 2.20, output 15.00 -> 13.20 (per million tokens)

cache_read_input_per_million (0.1x input) and cache_write_input_per_million
(1.25x input, per DUX-11232) are recomputed from the new input rates to
keep the same ratios that were already in place:

  Luna:  cache_read 0.1 -> 0.022, cache_write 1.25 -> 0.275
  Terra: cache_read 0.25 -> 0.22, cache_write 3.125 -> 2.75

Sol and all other models are untouched - this price cut only applies to
Luna/Terra.

Sources: AWS Bedrock pricing page (aws.amazon.com/bedrock/pricing) and
OpenAI's July 30, 2026 pricing announcement.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>

Co-authored-by: Sam Boland <sam.boland@appfolio.com>
---
 lib/ruby_llm/models.json                     | 32 ++++++++++----------
 spec/ruby_llm/models_gpt_5_6_bedrock_spec.rb | 10 ++++--
 2 files changed, 24 insertions(+), 18 deletions(-)

diff --git a/lib/ruby_llm/models.json b/lib/ruby_llm/models.json
index 9d983118d..28b241b02 100644
--- a/lib/ruby_llm/models.json
+++ b/lib/ruby_llm/models.json
@@ -16812,10 +16812,10 @@
     "pricing": {
       "text_tokens": {
         "standard": {
-          "input_per_million": 1.0,
-          "output_per_million": 6.0,
-          "cache_read_input_per_million": 0.1,
-          "cache_write_input_per_million": 1.25
+          "input_per_million": 0.22,
+          "output_per_million": 1.32,
+          "cache_read_input_per_million": 0.022,
+          "cache_write_input_per_million": 0.275
         }
       }
     },
@@ -16840,10 +16840,10 @@
         }
       ],
       "cost": {
-        "input": 1.0,
-        "output": 6.0,
-        "cache_read": 0.1,
-        "cache_write": 1.25
+        "input": 0.22,
+        "output": 1.32,
+        "cache_read": 0.022,
+        "cache_write": 0.275
       },
       "limit": {
         "context": 272000,
@@ -16948,10 +16948,10 @@
     "pricing": {
       "text_tokens": {
         "standard": {
-          "input_per_million": 2.5,
-          "output_per_million": 15.0,
-          "cache_read_input_per_million": 0.25,
-          "cache_write_input_per_million": 3.125
+          "input_per_million": 2.2,
+          "output_per_million": 13.2,
+          "cache_read_input_per_million": 0.22,
+          "cache_write_input_per_million": 2.75
         }
       }
     },
@@ -16976,10 +16976,10 @@
         }
       ],
       "cost": {
-        "input": 2.5,
-        "output": 15.0,
-        "cache_read": 0.25,
-        "cache_write": 3.125
+        "input": 2.2,
+        "output": 13.2,
+        "cache_read": 0.22,
+        "cache_write": 2.75
       },
       "limit": {
         "context": 272000,
diff --git a/spec/ruby_llm/models_gpt_5_6_bedrock_spec.rb b/spec/ruby_llm/models_gpt_5_6_bedrock_spec.rb
index 88bfd36b0..4475fa8ad 100644
--- a/spec/ruby_llm/models_gpt_5_6_bedrock_spec.rb
+++ b/spec/ruby_llm/models_gpt_5_6_bedrock_spec.rb
@@ -14,6 +14,12 @@
   # registry carrying a cache_write_input_per_million value — every other
   # OpenAI entry has none because the rate really is $0 for those models.
   #
+  # Luna/Terra base input/output rates reflect the July 30, 2026 OpenAI/Bedrock
+  # price cut (AWS Bedrock pricing page: aws.amazon.com/bedrock/pricing; OpenAI's
+  # July 30, 2026 pricing announcement). cache_read (0.1x) and cache_write (1.25x)
+  # remain the same ratios of the (now lower) input rate as before. Sol was not
+  # part of this price cut and is unchanged.
+  #
   # Separately open (do NOT resolve here): OpenAI's developer community has
   # reported two GPT-5.6-specific usage-accounting bugs since launch (July
   # 2026) — one where cached_tokens + cache_write_tokens could nearly
@@ -27,8 +33,8 @@
   # Luna/Terra/Sol costs as an open risk until this is checked.
   {
     'openai.gpt-5.6-sol' => { input: 5.0, output: 30.0, cache_read: 0.5, cache_write: 6.25 },
-    'openai.gpt-5.6-terra' => { input: 2.5, output: 15.0, cache_read: 0.25, cache_write: 3.125 },
-    'openai.gpt-5.6-luna' => { input: 1.0, output: 6.0, cache_read: 0.1, cache_write: 1.25 }
+    'openai.gpt-5.6-terra' => { input: 2.2, output: 13.2, cache_read: 0.22, cache_write: 2.75 },
+    'openai.gpt-5.6-luna' => { input: 0.22, output: 1.32, cache_read: 0.022, cache_write: 0.275 }
   }.each do |id, cost|
     it "resolves #{id} from the bedrock provider with the documented effort values" do
       model = RubyLLM.models.find(id, :bedrock)