How GPT-5.6 fuses frontier intelligence with frontier

What happened

6 model family to balance capability and cost across the spectrum of tasks people use our models for. 6 Sol, with max reasoning outperforms Claude Fable 5 on the Artificial Analysis Coding Agent Index at less than half of the cost. 5 on intelligence benchmarks at half the price, and Luna is our fastest and most affordable model, priced 80% less than the cost of Sol.

Globally, we route requests based on factors such as geography, available capacity, and accelerator type (the type of GPU or specialized chip running the model). Within a cluster, we distribute work across model instances based on load, context length, cache availability, and other request properties. Within each instance, work must then be partitioned efficiently across accelerators, the model’s sub-networks, and computing cores.

6 Sol in Codex helps us analyze production traffic, identify previously overlooked sources of imbalance, test new routing strategies, and constantly tune these heuristics. These load balancing improvements alone dramatically reduced the cost of serving our models. 6 Sol to optimize the model’s forward pass: the computation that transforms inputs into next-token predictions.

Even when individual operations are fast, excess memory movement, synchronization, and inefficient data layouts can leave GPUs idle. 6 Sol found work that could be precomputed, avoided, or parallelized. 6 Sol autonomously rewrote and optimized our production kernels, the core code that executes the mathematical operations that make up the model.

6 to be effective at writing and improving kernels in Triton⁠(opens in a new window) and Gluon⁠(opens in a new window), two open-source GPU programming languages maintained by OpenAI. 6 Sol, reduced end-to-end serving costs by 20%. 6 Sol. Speculative decoding is another lever for improving speed and efficiency.

Why it matters

To deliver these efficiencies, our research and technical teams have made significant optimizations at every major layer of our stack. These improvements span our models, inference (how we run models to generate output), and our agentic harness, which is used by both Codex and ChatGPT Work.

As we’ve scaled our models to 1 billion active users and more than 2 million businesses over the past four years, efficiency has been central to distributing the benefits of intelligence to everyone. Our mission is to ensure that artificial general intelligence benefits all of humanity.

Over these years, we’ve worked to continuously unlock greater optimizations across our stack in order to offer the most performant models at every point in the cost-intelligence curve. 6, which is trained to achieve more work per token. In training, we optimize for both task success and efficiency, shaping the model to take a more direct path through a task.

This post looks beyond our models to share how we designed for efficiency through advancements in two other major parts of the stack, including 1) inference, by optimizing processes such as load balancing, speculative decoding, caching, and kernel optimization, to get more output from the same hardware, and 2) our agentic harness, including better managing context bloat, tool usage, and repeated work.

6 Sol in landing several of these gains autonomously. While any isolated improvement may seem limited, these wins compound to allow us to deliver on the frontier of both intelligence and efficiency. In a compute-constrained world where model demand is growing faster than capacity, efficiency is core to every system design. That’s especially true in our inference stack, which runs trained models to generate responses.

Our primary objective is to serve more tokens with the same hardware, while preserving the intelligence, latency, availability, and reliability users expect. Achieving this requires optimizing the entire system. A model can be highly efficient in isolation, but still be expensive to serve if requests are distributed poorly, hardware sits idle, or data movement slows down computation.

Improvements at every layer compound, with gains coming from optimizations in routing (where requests are sent), scheduling (when requests are sent), kernels (software that runs on GPUs), caching (saved and reused work), and model implementation (the ordering of GPU code). 6 Sol in Codex played an instrumental role in all of these optimizations. The first important example is load balancing.

What to watch

The technique involves running a smaller draft (or “speculator”) model alongside the primary model, proposing several tokens for the primary model to verify in parallel. When those proposals are accepted, the system can produce multiple output tokens from a single primary-model pass, reducing the amount of expensive sequential computation.

6 Sol improved its own draft model by designing and running hundreds of experiments on its architecture, testing changes in size, structure, and features.