# GPT-6 Astra

Variational quantum algorithms normally rely on a classical optimizer to tune gate angles, and every loss evaluation is a run of the quantum device. A study has the frontier model GPT-6 Astra write the angles directly from the problem, circuit and gate conventions, with no loss or gradient returned; on transverse-field Ising chains, H2 and depth-one QAOA its angles match the optimizer, and as a warm start they cut the median optimizer evaluations to best value down to one.

Variational quantum algorithms normally rely on a classical optimizer to tune gate angles, and every loss evaluation is a run of the quantum device. A study has the frontier model GPT-6 Astra write the angles directly from the problem, circuit and gate conventions, with no loss or gradient returned; on transverse-field Ising chains, H2 and depth-one QAOA its angles match the optimizer, and as a warm start they cut the median optimizer evaluations to best value down to one.

Clinical prediction models often arrive as files without the data they were trained on, so an auditor must work out what the model actually computes. Researchers trained tanh networks of 105 to 38,401 weights on simulated subarachnoid haemorrhage cohorts, planted known relationships or a label bias, and asked GPT-6 Astra to answer with no tools, with a calculation-only white-box tool, or with a black-box tool exposing only predictions. Without tools Astra recovered planted relationships in 353-weight networks but became unreliable at 3,457 weights; with the calculator it recovered the relationship in all 29 runs that had one, raised no false alarm in 15 null runs, and black-box matched white-box.

Clinical prediction models often arrive as files without the data they were trained on, so an auditor must work out what the model actually computes. Researchers trained tanh networks of 105 to 38,401 weights on simulated subarachnoid haemorrhage cohorts, planted known relationships or a label bias, and asked GPT-6 Astra to answer with no tools, with a calculation-only white-box tool, or with a black-box tool exposing only predictions. Without tools Astra recovered planted relationships in 353-weight networks but became unreliable at 3,457 weights; with the calculator it recovered the relationship in all 29 runs that had one, raised no false alarm in 15 null runs, and black-box matched white-box.

OpenAI reports model progress on low-resource programming languages such as COBOL, Fortran and Visual Basic, with two evaluations: MultiLangBench asks a model to build a SQLite-compatible database engine from scratch, and a COBOL migration evaluation asks it to migrate programs while preserving behavior. On MultiLangBench, GPT-6 Astra averages an 87.4% test pass rate across 20 languages versus 26.8% for GPT-5.6 Sol; on COBOL migration, GPT-6.1 Sol reaches 86.6% at low reasoning effort versus 5.4%.

OpenAI reports model progress on low-resource programming languages such as COBOL, Fortran and Visual Basic, with two evaluations: MultiLangBench asks a model to build a SQLite-compatible database engine from scratch, and a COBOL migration evaluation asks it to migrate programs while preserving behavior. On MultiLangBench, GPT-6 Astra averages an 87.4% test pass rate across 20 languages versus 26.8% for GPT-5.6 Sol; on COBOL migration, GPT-6.1 Sol reaches 86.6% at low reasoning effort versus 5.4%.

OpenAI partnered with AI contracting platform Ironclad to turn legal, commercial, and procurement contracting workflows into 11 research tasks scored against 8 to 50 criteria each, training models with synthetic tasks and reinforcement learning inside Ironclad-hosted software environments; across the 11 tasks GPT‑6 Astra averaged 55.0% versus 41.6% for GPT‑5.6 Sol, with estimated average time per attempt falling from 37.0 to 19.2 minutes.

OpenAI partnered with AI contracting platform Ironclad to turn legal, commercial, and procurement contracting workflows into 11 research tasks scored against 8 to 50 criteria each, training models with synthetic tasks and reinforcement learning inside Ironclad-hosted software environments; across the 11 tasks GPT‑6 Astra averaged 55.0% versus 41.6% for GPT‑5.6 Sol, with estimated average time per attempt falling from 37.0 to 19.2 minutes.

PyRUA-Lean is an interactive code-execution framework in which a VLM agent composes classical robot primitives and frozen VLA policies into Python cells that perform conditional checks and local retries and return only explicitly requested images and state feedback; across 700 simulated task instances from LIBERO-PRO, RoboTwin 2.0, and RoboCasa365, and against a tool-calling baseline using the same GPT-6 Astra planner and the same primitives, it raises overall success from 63.1% to 71.7% under equal LLM-call budgets and uses 49% fewer LLM calls and 65% fewer input tokens on instances solved by both agents.

PyRUA-Lean is an interactive code-execution framework in which a VLM agent composes classical robot primitives and frozen VLA policies into Python cells that perform conditional checks and local retries and return only explicitly requested images and state feedback; across 700 simulated task instances from LIBERO-PRO, RoboTwin 2.0, and RoboCasa365, and against a tool-calling baseline using the same GPT-6 Astra planner and the same primitives, it raises overall success from 63.1% to 71.7% under equal LLM-call budgets and uses 49% fewer LLM calls and 65% fewer input tokens on instances solved by both agents.

This work systematically evaluates GPT-6 Astra as an embodied policy across six domains—gripper manipulation, dexterous manipulation, mobile manipulation, navigation, locomotion, and humanoid loco-manipulation—comparing direct control with hybrid control that cooperates with learned policies or whole-body controllers, finding leading navigation results (92% on RxR instruction following, 82% on HM3D object search), hybrid success of 48% on RoboDojo, 50% on DexJoCo, and 38.7% on RoboCasa365, while direct in-hand control and dense-reference locomotion remain unreliable, with substantial token use and inference latency recorded.

This work systematically evaluates GPT-6 Astra as an embodied policy across six domains—gripper manipulation, dexterous manipulation, mobile manipulation, navigation, locomotion, and humanoid loco-manipulation—comparing direct control with hybrid control that cooperates with learned policies or whole-body controllers, finding leading navigation results (92% on RxR instruction following, 82% on HM3D object search), hybrid success of 48% on RoboDojo, 50% on DexJoCo, and 38.7% on RoboCasa365, while direct in-hand control and dense-reference locomotion remain unreliable, with substantial token use and inference latency recorded.

At DevDay 2026, OpenAI announced more than 20 items spanning GPT-6 Astra, ChatGPT, Codex, APIs, security, and new tools for builders.

At DevDay 2026, OpenAI announced more than 20 items spanning GPT-6 Astra, ChatGPT, Codex, APIs, security, and new tools for builders.

Researchers at MBZUAI and Apertix placed GPT-6 Astra alongside Fable 5, Kimi K3, Gemini 3.1 Pro, Qwen 3.8-Max, and Muse Spark 1.3 under identical instructions, inputs, and scoring across 34 capabilities, 55 benchmarks, and nine areas of computer vision, comparing them with dedicated models and humans, and found that semantic interpretation, reasoning, and object-centric prediction approach or reach reference levels while metric geometric accuracy, faithful reconstruction, temporally consistent dense prediction, and specialized fine-grained visual knowledge retain larger gaps.

Researchers at MBZUAI and Apertix placed GPT-6 Astra alongside Fable 5, Kimi K3, Gemini 3.1 Pro, Qwen 3.8-Max, and Muse Spark 1.3 under identical instructions, inputs, and scoring across 34 capabilities, 55 benchmarks, and nine areas of computer vision, comparing them with dedicated models and humans, and found that semantic interpretation, reasoning, and object-centric prediction approach or reach reference levels while metric geometric accuracy, faithful reconstruction, temporally consistent dense prediction, and specialized fine-grained visual knowledge retain larger gaps.