GPT-5.6 Sol vs GPT-6 Astra
Terminal-Bench 4.0 Terminal-Bench 4.0 · source setting 2: Agents use a terminal to complete executable coding, system administration and research tasks in the 4.0 suite
GPT-5.6 Sol14.6
GPT-6 Astra49.5
DeepSWE 1.1 DeepSWE 1.1: Agents change repository code to resolve software engineering tasks under the stated DeepSWE edition
GPT-5.6 Sol61.1
GPT-6 Astra72.8
SWE-bench Pro (v2 full) SWE-bench Pro (v2 full): Agents resolve repository issues under the network-locked v2 protocol; full and 51-task Hard results are separate
GPT-5.6 Sol · xhigh95.5
GPT-6 Astra · high96.9
SWE-bench Pro (v2 hard) SWE-bench Pro (v2 hard): Agents resolve repository issues under the network-locked v2 protocol; full and 51-task Hard results are separate
GPT-5.6 Sol · xhigh82.4
GPT-6 Astra · high90.2
SciCode SciCode: Writing code for real scientific computing problems (physics, biology, math)
GPT-5.6 Sol57.4
GPT-6 Astra54.2
Terminal-Bench Science 0.1 Terminal-Bench Science 0.1: Agents complete expert-authored research workflows across five scientific domains in a terminal
GPT-5.6 Sol · max22.4
GPT-6 Astra · max63.3
MultiLangBenchlab MultiLangBench: Models build a SQLite-compatible database engine in 20 programming languages
GPT-5.6 Sol26.8
GPT-6 Astra87.4
MirrorCode MirrorCode: Agents reimplement software projects by observing their behavior
GPT-5.6 Sol · high20
GPT-6 Astra · high46.7
FrontierCode 1.1 FrontierCode 1.1: Agents implement changes in real repositories, graded against maintainer-defined requirements
GPT-5.6 Sol—
GPT-6 Astra · max47
WebDev Arena WebDev Arena: Human pairwise preferences in the stated webdev vote pool
GPT-5.6 Sol—
GPT-6 Astra · max1,774
Terminal-Bench Hard (version unstated) Terminal-Bench Hard (version unstated): Agents solve difficult software, system-administration and data-processing tasks in a terminal
GPT-5.6 Sol62.9
GPT-6 Astra—
| Benchmark | GPT-5.6 Sol | GPT-6 Astra |
|---|---|---|
| Terminal-Bench 4.0 | 14.6 | 49.5 |
| DeepSWE 1.1 | 61.1 | 72.8 |
| SWE-bench Pro (v2 full) | 95.5 · xhigh | 96.9 · high |
| SWE-bench Pro (v2 hard) | 82.4 · xhigh | 90.2 · high |
| SciCode | 57.4 | 54.2 |
| Terminal-Bench Science 0.1 | 22.4 · max | 63.3 · max |
| MultiLangBenchlab | 26.8 | 87.4 |
| MirrorCode | 20 · high | 46.7 · high |
| FrontierCode 1.1 | — | 47 · max |
| WebDev Arena | — | 1,774 · max |
| Terminal-Bench Hard (version unstated) | 62.9 | — |
Specs
| Spec | GPT-5.6 Sol | GPT-6 Astra |
|---|---|---|
| Input, per 1M | $4 | $10 |
| Output, per 1M | $20 | $50 |
| Cost per task? | $0.50 | $1.54 |
| Speed, tok/s | 85 | 59 |
| Context | 1.05M | 1.05M |
| Weights | Closed | Closed |
| Released | Jul 2026 | Sep 2026 |
| See images | ✓ | ✓ |
| Read PDFs | Partial | Partial |
| Hear audio | — | — |
| Watch video | — | — |
| Call tools | ✓ | ✓ |
| Force a tool call | Partial | Partial |
| Return strict JSON | ✓ | ✓ |
| Search the web | ? | ✓ |
| Run code | ? | ✓ |
| Use a computer | ? | ✓ |
What breaks when you switch
GPT-5.6 Sol → GPT-6 Astra
Monitoring can return HTTP 403›
Misalignment monitoring can return HTTP 403 before streaming or errors after output; match misalignment_policy_violation and handle stream errors.
No none reasoning effort›
The none reasoning effort is unsupported on GPT-6 Astra and GPT-6.1 Sol; use low instead.
Remove unsupported sampling parameters›
When reasoning effort is not none, remove temperature, top_p and top_logprobs, plus Chat Completions logprobs or Responses output logprobs inclusion.
GPT-6 Astra → GPT-5.6 Sol
Nothing known.