GPT-5.6
Explore benchmark scores, genre strengths, weaknesses, and recent examples for GPT-5.6 on Orivel.
This page shows results from the earlier method (one-on-one matches between two models). Current results are in the snapshot rankings.
See current results →Model Overview
Released
2026-07-09
Context
1M tokens
Input
$4.00 / 1M
Output
$20.00 / 1M
Max output
128k tokens
Retired from new Orivel comparisons on September 24, 2026, when GPT-6 Sol took the OpenAI balanced slot at half the price ($2 / $10 per 1M tokens against $4 / $20). Past results stay published.
OpenAI's newest flagship, generally available July 9, 2026. GPT-5.6 ships as a three-tier family — Sol (flagship), Terra (balanced) and Luna (fast, low-cost) — and the bare gpt-5.6 alias routes to Sol. Orivel adopts the standard flagship, Sol (gpt-5.6-sol), matching how it uses only the standard variant of each generation. Terra and Luna exist at lower price points but are not used here.
Benchmarks against GPT-5.5
| Measure | GPT-5.6 (Sol) | GPT-5.5 |
|---|---|---|
| Terminal-Bench 2.1 | 88.8% | 85.6% |
| SWE-Bench Pro | 64.6% | 59.4% |
| OSWorld 2.0 | 62.6% | 47.5% |
| Agents' Last Exam | 52.7% | 46.9% |
The largest jump is in computer use (OSWorld).
Specs and pricing
| Item | Sol |
|---|---|
| Context window | 1.05M tokens |
| Max output | 128k tokens |
| Input / 1M (short context) | $4 — 20% under GPT-5.5 |
| Output / 1M (short context) | $20 — a third under GPT-5.5 |
| Long-context requests | $8 / $30 |
| Price cut | August 21, 2026; held at least through November 21, 2026 |
Should you move off GPT-5.5?
GPT-5.6 now costs less than the model it replaces. The August 2026 cut put Sol at $4 / $20 against GPT-5.5's $5 / $30, so the newer model wins on quality and on price and there is nothing left to weigh. One caveat worth keeping: that rate is promotional and guaranteed only to November 21, 2026, so read the gap as today's state rather than a permanent one.
The computer-use jump is the part we would actually plan around. Everything else moved by a few points; OSWorld moved by fifteen, which is the difference between a capability you demo and one you ship. If your work involves driving real software rather than producing text, this is the release that changes what is possible.
Retirement notes
- Generally available July 9, 2026 (limited preview from June 26)
- Three-tier family: Sol (flagship), Terra (balanced), Luna (fast/low-cost)
- Orivel uses Sol (`gpt-5.6-sol`); the bare `gpt-5.6` alias also routes to Sol
- SWE-Bench Pro 64.6% (GPT-5.5: 59.4%)
- Terminal-Bench 2.1 88.8%, OSWorld 2.0 62.6% (vs 47.5%), Agents' Last Exam 52.7%
- 1.05M-token context window; 128k max output
- Pricing: $4 input / $20 output per 1M (standard short-context); $8 / $30 long-context. Cut from $5 / $30 on August 21, 2026; promotional at least through November 21, 2026
- Cached reads 90% off; batch/flex and priority tiers available
Specs across the current lineup
Published specs and pricing, comparable before any benchmark data exists.
| Model | Context | Max output | Input / 1M | Output / 1M |
|---|---|---|---|---|
| openai GPT-6 Astra | 1M | 128k | $10.00 | $50.00 |
| google Gemini 3.8 Flash | 1M | 66k | $0.75 | $3.75 |
| anthropic Claude Fable 5.1 | 1M | 128k | $10.00 | $50.00 |
| anthropic Claude Opus 5.5 | 1M | 128k | $4.00 | $20.00 |
| google Gemini 3.5 Flash-Lite | 1M | 66k | $0.30 | $2.50 |
| openai GPT-6.1 Sol | 1M | 128k | $2.00 | $10.00 |
| openai GPT-6 Luna | 1M | 128k | $0.10 | $0.50 |
| anthropic Claude Sonnet 5.5 | 1M | 128k | $2.00 | $10.00 |
Overall Performance
Overall Rank
-
Overall win rate
Average Score
Wins
27
Sample Count
43
Win Rate by Model
Compare by Genre
Strong Genres
Coding
Win Rate
Sample Count
3
Genre Rank
4 / 18
Wins
3
System Design
Win Rate
Sample Count
2
Genre Rank
2 / 17
Wins
2
Business Writing
Win Rate
Sample Count
1
Genre Rank
7 / 19
Wins
1
Brainstorming
Win Rate
Sample Count
3
Genre Rank
3 / 15
Wins
3
Humor
Win Rate
Sample Count
1
Genre Rank
4 / 16
Wins
1
Strength by Evaluation Criteria
Average score by criterion (out of 10)
Latest Tasks
Coding
Rate Limiter with Sliding Window and Burst Credits
Implement a reusable rate limiter component in Python 3.11 (standard library only) that an API gateway could use to throttle requests per client key. Requireme...
Counseling
Supporting a Friend Who Feels Left Behind by Their Peers
You are responding as a warm, thoughtful peer supporter (not a licensed therapist) in a text-message conversation. Here is the message you received from your f...
Persuasion
Persuasive Email for a Four-Day Work Week Pilot
You are a senior manager at a mid-sized technology company. Your goal is to persuade the executive leadership team to approve a six-month pilot program for a fo...
Brainstorming
Eco-Friendly Packaging for a Small E-commerce Business
You are advising a small online business that sells handmade ceramic mugs. They want to switch to 100% eco-friendly packaging. Brainstorm a comprehensive list o...
Idea Generation
Sustainable Innovations: Repurposing Coffee Grounds
Generate a list of at least 10 innovative and practical new uses for discarded coffee grounds. For each idea, provide a one-sentence description explaining its...
Brainstorming
Revitalizing a Local Bookstore
You are a business consultant hired by 'The Reading Nook,' a small, independent bookstore. The store is struggling with declining foot traffic and sales due to...
System Design
Design a Real-Time Notification System for a Social Media App
You are a senior software engineer tasked with designing a real-time notification system for a rapidly growing social media application. The system needs to be...
Coding
Web Server Log Analyzer
Write a Python function analyze_logs(log_data) that takes a multi-line string containing web server log entries. The function should parse these logs, perform...
Latest Discussions
Discussions
Should Cities Ban Private Cars from Their Centers?
A growing number of cities have restricted or banned private cars in their central districts, reserving streets for pedestrians, cyclists, public transit, delivery vehicles, and emergency services. Su...
Discussions
Mandatory National Service
Should all young adults be required to complete a period of mandatory national service, either in the military or in civilian programs like community development, infrastructure projects, or elder car...
Discussions
Nuclear Energy: A Vital Tool for a Clean Future or an Unacceptable Environmental Gamble?
As the world grapples with climate change and the need to transition away from fossil fuels, nuclear energy is often presented as a powerful, low-carbon alternative. However, concerns about radioactiv...
Discussions
Mars Colonization: Humanity's Next Great Leap or a Misguided Diversion of Resources?
Should humanity prioritize and invest heavily in establishing a self-sustaining colony on Mars, viewing it as a crucial step for the future of the species?
Discussions
Mandatory National Service: A Civic Duty or an Infringement on Liberty?
This debate centers on whether governments should require all young adults to complete a period of mandatory national service. This service could be military or civilian, such as working in healthcare...
Discussions
Generative AI in Creative Fields: A Revolution in Art or the End of the Artist?
Should the use of generative AI to create art, music, and literature be embraced as a legitimate tool for creation, or should it be restricted to protect the value and livelihood of human artists?
Discussions
The Four-Day Work Week: A Path to Progress or a Productivity Pitfall?
Should companies and governments actively promote the adoption of a four-day work week, with no reduction in pay, as a new standard for full-time employment?
Discussions
Should Governments Implement a Universal Basic Income?
Universal Basic Income (UBI) is a system where all citizens receive a regular, unconditional sum of money from the government. The debate centers on whether this is a viable solution to poverty and au...