Document GPT-5.6 models: comparison task sections, model cards, and cache write pricing (#62573)

Co-authored-by: Copilot <223556219+Copilot@users.noreply.github.com>
Copilot-Session: 3d652994-a326-4b43-a774-a276c9b8470d

Author: sunbrye

content/copilot/reference/ai-models/model-comparison.md

@@ -39,6 +39,7 @@ Use these models for common development tasks that require a balance of quality,
 |---------------------------------------------------|-------------------------------------------------------------------------------------------------------------------------------------------------|
 | {% data variables.copilot.copilot_gpt_53_codex %} | Delivers higher-quality code on complex engineering tasks like features, tests, debugging, refactors, and reviews without lengthy instructions. |
 | {% data variables.copilot.copilot_gpt_5_mini %}   | Reliable default for most coding and writing tasks. Fast, accurate, and works well across languages and frameworks.                             |
+| {% data variables.copilot.copilot_gpt_56_terra %} | Balanced all-round choice for everyday interactive and agentic coding.                                                                          |
 | {% data variables.copilot.copilot_raptor_mini %}  | Specialized for fast, accurate inline suggestions and explanations.                                        |
 | {% data variables.copilot.copilot_mai_code_1_flash %} | Strong instruction-following and adaptive reasoning make it a reliable default for everyday coding tasks, writing, and multi-turn development workflows. |
 
@@ -63,6 +64,7 @@ These models are optimized for speed and responsiveness. They’re ideal for qui
 
 | Model                                                 | Why it's a good fit                                                                                        |
 |-------------------------------------------------------|------------------------------------------------------------------------------------------------------------|
+| {% data variables.copilot.copilot_gpt_56_luna %}      | Lightweight, cost-efficient option for smaller, faster tasks. The lowest-cost model in the GPT-5.6 family. |
 | {% data variables.copilot.copilot_claude_haiku_45 %}  | Balances fast responses with quality output. Ideal for small tasks and lightweight code explanations.      |
 | {% data variables.copilot.copilot_mai_code_1_flash %} | Handles quick coding tasks with adaptive efficiency, stays concise for simple requests and delivers fast, accurate responses without unnecessary depth. |
 
@@ -90,6 +92,7 @@ These models are designed for tasks that require step-by-step reasoning, complex
 |-------------------------------------------------------|-----------------------------------------------------------------------------------------------------------------------------------------------------------------|
 | {% data variables.copilot.copilot_gpt_5_mini %}       | Delivers deep reasoning and debugging with faster responses and lower resource usage than GPT-5. Ideal for interactive sessions and step-by-step code analysis. |
 | {% data variables.copilot.copilot_gpt_55 %}           | Great at complex reasoning, code analysis, and technical decision-making.                                                                                       |
+| {% data variables.copilot.copilot_gpt_56_sol %}       | The highest reasoning ceiling in the GPT-5.6 family. Best for complex reasoning over large codebases and demanding, long-running agentic work. |
 | {% data variables.copilot.copilot_claude_sonnet_46 %} | Improves on Sonnet 4.5 with more reliable completions and smarter reasoning under pressure.                                                                     |
 | {% data variables.copilot.copilot_claude_opus_47 %}   | Anthropic’s most powerful model. Improves on {% data variables.copilot.copilot_claude_opus_46 %}.                                                               |
 | {% data variables.copilot.copilot_gemini_31_pro %}    | Advanced reasoning across long contexts and scientific or technical analysis.                                                                                   |

content/copilot/reference/copilot-billing/models-and-pricing.md

@@ -33,10 +33,12 @@ All prices are **per 1 million tokens**.
 
 {% data reusables.copilot.extended-context-pricing %}
 
-| Model | Release status | Category | Tier | Threshold (input tokens) | Input | Cached input | Output |
-| --- | --- | --- | --- | --- | ---: | ---: | ---: |
+{% data variables.copilot.copilot_gpt_56_sol %}, {% data variables.copilot.copilot_gpt_56_terra %}, and {% data variables.copilot.copilot_gpt_56_luna %} include a cache write cost in addition to cached input. Earlier OpenAI models have no cache write cost.
+
+| Model | Release status | Category | Tier | Threshold (input tokens) | Input | Cached input | Cache write | Output |
+| --- | --- | --- | --- | --- | ---: | ---: | ---: | ---: |
 | {% for entry in tables.copilot.models-and-pricing %}{% if entry.provider == "openai" %} |
-| {{ entry.model }} | {{ entry.release_status }} | {{ entry.category }} | {{ entry.tier }} | {{ entry.threshold }} | {{ entry.input }} | {{ entry.cached_input }} | {{ entry.output }} |
+| {{ entry.model }} | {{ entry.release_status }} | {{ entry.category }} | {{ entry.tier }} | {{ entry.threshold }} | {{ entry.input }} | {{ entry.cached_input }} | {{ entry.cache_write }} | {{ entry.output }} |
 | {% endif %}{% endfor %} |
 
 ### Anthropic

data/tables/copilot/model-comparison.yml

@@ -36,17 +36,17 @@
 - name: GPT-5.6 Luna
   task_area: Fast help with simple or repetitive tasks
   excels_at: Quick, cost-efficient responses for smaller, faster coding tasks
-  further_reading: 'Not available'
+  further_reading: '[GPT-5.6 model card](https://deploymentsafety.openai.com/gpt-5-6/gpt-5-6.pdf)'
 
 - name: GPT-5.6 Sol
   task_area: Deep reasoning and debugging
   excels_at: Complex reasoning over large codebases and long-running agentic work
-  further_reading: 'Not available'
+  further_reading: '[GPT-5.6 model card](https://deploymentsafety.openai.com/gpt-5-6/gpt-5-6.pdf)'
 
 - name: GPT-5.6 Terra
   task_area: General-purpose coding and agent tasks
   excels_at: Balanced everyday interactive and agentic coding
-  further_reading: 'Not available'
+  further_reading: '[GPT-5.6 model card](https://deploymentsafety.openai.com/gpt-5-6/gpt-5-6.pdf)'
 
 # Anthropic
 - name: Claude Fable 5

data/tables/copilot/models-and-pricing.yml

@@ -14,7 +14,9 @@
 # - output: Output token price per 1M tokens.
 # - threshold: Input token threshold for pricing tier (OpenAI and Google only).
 # - tier: Pricing tier label (OpenAI and Google only).
-# - cache_write: Cache write price per 1M tokens (Anthropic only).
+# - cache_write: Cache write price per 1M tokens. Applies to Anthropic models and
+#   to GPT-5.6 and later OpenAI model families. Use "Not applicable" for models
+#   that have no cache write cost.
 # - notes: Optional notes about the model.
 
 # OpenAI
@@ -27,6 +29,7 @@
   input: $0.25
   cached_input: $0.025
   output: $2.00
+  cache_write: Not applicable
 
 - model: GPT-5.3-Codex
   provider: openai
@@ -37,6 +40,7 @@
   input: $1.75
   cached_input: $0.175
   output: $14.00
+  cache_write: Not applicable
 
 - model: 'GPT-5.4'
   provider: openai
@@ -47,6 +51,7 @@
   input: $2.50
   cached_input: $0.25
   output: $15.00
+  cache_write: Not applicable
 
 - model: GPT-5.4
   provider: openai
@@ -57,6 +62,7 @@
   input: $5.00
   cached_input: $0.50
   output: $22.50
+  cache_write: Not applicable
 
 - model: GPT-5.4 mini
   provider: openai
@@ -67,6 +73,7 @@
   input: $0.75
   cached_input: $0.075
   output: $4.50
+  cache_write: Not applicable
 
 - model: GPT-5.4 nano
   provider: openai
@@ -77,6 +84,7 @@
   input: $0.20
   cached_input: $0.02
   output: $1.25
+  cache_write: Not applicable
 
 - model: GPT-5.5
   provider: openai
@@ -87,6 +95,7 @@
   input: $5.00
   cached_input: $0.50
   output: $30.00
+  cache_write: Not applicable
 
 - model: GPT-5.5
   provider: openai
@@ -97,6 +106,7 @@
   input: $10.00
   cached_input: $1.00
   output: $45.00
+  cache_write: Not applicable
 
 - model: GPT-5.6 Luna
   provider: openai
@@ -107,6 +117,7 @@
   input: $0.20
   cached_input: $0.02
   output: $1.20
+  cache_write: $0.25
 
 - model: GPT-5.6 Luna
   provider: openai
@@ -117,6 +128,7 @@
   input: $0.40
   cached_input: $0.04
   output: $1.80
+  cache_write: $0.50
 
 - model: GPT-5.6 Sol
   provider: openai
@@ -127,6 +139,7 @@
   input: $5.00
   cached_input: $0.50
   output: $30.00
+  cache_write: $6.25
 
 - model: GPT-5.6 Sol
   provider: openai
@@ -137,6 +150,7 @@
   input: $10.00
   cached_input: $1.00
   output: $45.00
+  cache_write: $12.50
 
 - model: GPT-5.6 Terra
   provider: openai
@@ -147,6 +161,7 @@
   input: $2.00
   cached_input: $0.20
   output: $12.00
+  cache_write: $2.50
 
 - model: GPT-5.6 Terra
   provider: openai
@@ -157,6 +172,7 @@
   input: $4.00
   cached_input: $0.40
   output: $18.00
+  cache_write: $5.00
 
 # Anthropic
 - model: Claude Haiku 4.5