GPT-5.1 Codex Mini vs GPT-5.6 Terra
Different efforts: high · medium · Compare all at high
Humanity's Last Exam Humanity's Last Exam: Very hard expert questions across many fields, written to stump today’s models
GPT-5.1 Codex Mini18.5
GPT-5.6 Terra33.3
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-5.1 Codex Mini—
GPT-5.6 Terra1
AutomationBench-AA AutomationBench-AA: Agents orchestrate SaaS workflows through APIs; objectives receive credit only when the task has no guardrail violation
GPT-5.1 Codex Mini—
GPT-5.6 Terra33.5
Arena Text · creative writing Arena Text · creative writing: Human pairwise preferences in the stated text vote pool
GPT-5.1 Codex Mini—
GPT-5.6 Terra · xhigh1,429
MMMU-Pro (vision) MMMU-Pro (vision): College-level questions that require understanding images: charts, diagrams, photos
GPT-5.1 Codex Mini69
GPT-5.6 Terra76.8
| Benchmark | GPT-5.1 Codex Mini | GPT-5.6 Terra |
|---|---|---|
| Humanity's Last Exam | 18.5 | 33.3 |
| Terminal-Bench 4.0 | — | 1 |
| AutomationBench-AA | — | 33.5 |
| Arena Text · creative writing | — | 1,429 · xhigh |
| MMMU-Pro (vision) | 69 | 76.8 |
Specs
| Spec | GPT-5.1 Codex Mini | GPT-5.6 Terra |
|---|---|---|
| Input, per 1M | $0.25 | $2 |
| Output, per 1M | $2 | $12 |
| Cost per task? | — | $0.18 |
| Speed, tok/s | 172 | 125 |
| Context | 400K | 1.05M |
| Weights | Closed | Closed |
| Released | Nov 2025 | Jul 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-5.1 Codex Mini → GPT-5.6 Terra
Prompts with >272K input tokens›
Above 272K input tokens, input costs 2× and output 1.5× for the full request.
GPT-5.6 Terra → GPT-5.1 Codex Mini