GPT-5.6 Terra
No shutdown before this date. GPT-5.6 Terra stays on Amazon Bedrock until at least Jul 13, 2027.
≥ 9 mo- Earliest retirement
- ≥ 272 daysJul 13, 2027 · Amazon Bedrock
- Arena Text
- 1,466#27 of 128 · Arena
- GPQA Diamond
- 93.3%#16 of 108 · Epoch AI
- List price · $ per 1M
- $2.00 in $12.00 outOpenAI API · cached input $0.20
- Context window
- 1.1MMax output 128K tokens
Dates on every platform
| Platform | Model ID and status | Deprecated | Shuts down | Replacement |
|---|---|---|---|---|
| OpenAI APIOpenAI: List models (GET /v1/models) (opens in a new tab) · checked Oct 14, 2026 | gpt-5.6-terraActive | – | None set | – |
| Azure AI FoundryMicrosoft: Model retirement schedule - Microsoft Foundry (opens in a new tab) · checked Oct 14, 2026 | gpt-5.6-terra:2026-07-09Active | – | 15 mo | – |
| Amazon BedrockAmazon Web Services: Amazon Bedrock ListFoundationModels, us-east-1 (opens in a new tab) · checked Oct 14, 2026 | openai.gpt-5.6-terraActive · also global.openai.gpt-5.6-terra, in.openai.gpt-5.6-terra, us.openai.gpt-5.6-terra | – | ≥ 9 moNot sooner than | – |
In us-east-1, requests go through a cross-Region inference profile (listed in aliases).
What breaks when you move off GPT-5.6 Terra
- Defaults
- Pricing
Intelligence
| Benchmark | Publisher | Score | Rank | Published |
|---|---|---|---|---|
| Arena Text | Arena | 1,466 ± 2 | #27 of 128 | Oct 8, 2026 |
| GPQA Diamond | Epoch AI | 93.3% ± 1.5% | #16 of 108 | Jul 9, 2026 |
| Arena Agent | Arena | 0 ± 0 | #23 of 35 | Oct 12, 2026 |
| FrontierMath Tier 4 | Epoch AI | 70.7% ± 7.2% | #12 of 58 | Jul 9, 2026 |
| FrontierMath Tiers 1-3 | Epoch AI | 86.0% ± 2.1% | #10 of 71 | Jul 9, 2026 |
| OTIS Mock AIME 2024-2025 | Epoch AI | 99.7% ± 0.3% | #9 of 105 | Jul 9, 2026 |
| SimpleQA Verified | Epoch AI | 43.2% ± 1.6% | #35 of 68 | Aug 10, 2026 |
Arena (arena.ai), https://arena.ai/leaderboard
Epoch AI, ‘Capabilities & benchmarking’. Published online at epoch.ai. Retrieved from ‘https://epoch.ai/benchmarks’ [online resource].
List prices
| Platform | Input | Cached | Output | In effect | Source |
|---|---|---|---|---|---|
OpenAI APIgpt-5.6-terra | $2.00 | $0.20 | $12.00 | Jul 30, 2026Current · Requests over 272K input tokens are billed at higher long-context rates. Cache writes are billed separately. Batch and Flex bill less; Fast mode bills more. | OpenAI: Pricing (opens in a new tab) |
Azure AI Foundrygpt-5.6-terra:2026-07-09 | $2.00 | $0.20 | $12.00 | Aug 1, 2026Current · Global Standard short-context rate from the US East 2 price list. | Microsoft: Azure Retail Prices API (Foundry Models, US East 2) (opens in a new tab) |
Amazon Bedrockopenai.gpt-5.6-terra | $2.20 | $0.22 | $13.20 | Oct 14, 2026Current · Standard tier, short context (272K input tokens or fewer). In-Region and geographic cross-Region inference; global cross-Region inference is priced lower. | Amazon Web Services: Amazon Bedrock model card: GPT-5.6 Terra (opens in a new tab) |
Change history
ShutdownGPT-3.5 Turbo Instruct, babbage-002, davinci-002 and gpt-3.5-turbo-1106 shut downOpenAI API1 break
OpenAI removed gpt-3.5-turbo-instruct, babbage-002, davinci-002 and the gpt-3.5-turbo-1106 snapshot. It recommends GPT-5.6 Terra.
- Platform
- OpenAI API
- Replacement
- GPT-5.6 Terra
- Announced
- Toolingbabbage-002 and davinci-002 used the legacy Completions endpoint. GPT-5.6 Terra does not support v1/completions.
Source: OpenAI, Deprecations (opens in a new tab); OpenAI, GPT-5.6 Terra model page (opens in a new tab)
ReleaseGPT-5.6 Luna, Sol and Terra available on Amazon BedrockAmazon Bedrock
Bedrock added GPT-5.6 Luna, GPT-5.6 Sol and GPT-5.6 Terra.
- Platform
- Amazon Bedrock
- Models
- GPT-5.6 Luna, GPT-5.6 Sol, GPT-5.6 Terra
Source: Amazon Web Services, Amazon Bedrock ListFoundationModels, us-east-1 (opens in a new tab)
Price changeGPT-5.6 Luna and Terra prices cut on the OpenAI APIOpenAI API
GPT-5.6 Luna costs 80 percent less and GPT-5.6 Terra 20 percent less from 30 July 2026. Fast mode replaced Priority processing the same day; requests tagged priority use Fast mode.
- Platform
- OpenAI API
- Models
- GPT-5.6 Luna, GPT-5.6 Terra
ReleaseGPT-5.6 Sol, Terra and Luna available on Azure AI FoundryAzure AI Foundry
Microsoft made the GPT-5.6 series generally available in Foundry Models and Foundry Agent Service, with Global Standard, Global Priority Processing, Data Zone Standard and Global Provisioned deployments from day one.
- Platform
- Azure AI Foundry
- Models
- GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna
Source: Microsoft, GPT-5.6 now available in Microsoft Foundry (opens in a new tab)
ReleaseGPT-5.6 Sol, Terra and Luna released on the OpenAI APIOpenAI API2 breaks
OpenAI released the GPT-5.6 family: Sol for frontier capability, Terra for a balance of intelligence and cost, and Luna for high-volume work. The gpt-5.6 alias routes to gpt-5.6-sol. The family adds programmatic tool calling, explicit prompt caching controls, persisted reasoning, max reasoning effort and pro mode.
- Platform
- OpenAI API
- Models
- GPT-5.6 Sol, GPT-5.6 Terra, GPT-5.6 Luna
- DefaultsGPT-5.6 models reuse reasoning across turns by default (reasoning.context all_turns); earlier models default to current_turn.
- PricingGPT-5.6 bills cache writes at 1.25 times the uncached input rate. Replace prompt_cache_retention with prompt_cache_options.ttl.
Source: OpenAI, Changelog (opens in a new tab); OpenAI, Using GPT-5.6 (opens in a new tab)