Model comparison
GPT-5.3 Codex vs GPT-5.6 Terra
Head-to-head evidence from 14 shared benchmark results across 6 categories. Overall scores shown here use the public BenchAlign v5 ranking lane.
Public leaderboard positions: GPT-5.3 Codex #26 (Supported); GPT-5.6 Terra #11 (Estimated). Intervals and evidence labels describe ranking uncertainty, not a guarantee for a specific workload.
Evidence parity. GPT-5.3 Codex and GPT-5.6 Terra share 14 comparable benchmark results. 2 of 8 categories are comparable. 7 results are unique to GPT-5.3 Codex; 30 to GPT-5.6 Terra.
Updated July 23, 2026- Shared results
- 14
- GPT-5.3 Codex only
- 7
- GPT-5.6 Terra only
- 30
- Comparable categories
- 2 / 8
Pick GPT-5.6 Terra if you want the stronger benchmark profile. GPT-5.3 Codex only becomes the better choice if coding is the priority or you want the cheaper token bill.
Confidence note. This is a partial-evidence comparison with 14 shared benchmark results across 6 evidence categories; 2 of 8 categories currently have scoreable aggregates for both models. Treat the verdict as directional until coverage is more balanced.
Why this result
GPT-5.6 Terra is clearly ahead on the BenchAlign aggregate, 72.57 to 66.69. The gap is large enough that you do not need to squint at the spreadsheet to see the difference.
GPT-5.6 Terra's sharpest advantage is in agentic, where it averages 87.4 against 71.4. The single biggest benchmark swing on the page is Terminal-Bench 2.0, 77.3% to 87.4%. GPT-5.3 Codex does hit back in coding, so the answer changes if that is the part of the workload you care about most.
GPT-5.6 Terra is also the more expensive model on tokens at $2.50 input / $15.00 output per 1M tokens, versus $1.75 input / $14.00 output per 1M tokens for GPT-5.3 Codex. GPT-5.6 Terra gives you the larger context window at 1M, compared with 400K for GPT-5.3 Codex.
Category breakdown
Exact category averages are shown below. Not measured means BenchLM does not have enough sourced public coverage for that model and category.
| Category | GPT-5.3 Codex | Δ | GPT-5.6 Terra |
|---|---|---|---|
| Agentic | GPT-5.3 Codex71.4 | Margin→ 16.0 | GPT-5.6 Terra87.4 |
| Coding | GPT-5.3 Codex67.2 | Margin← 3.8 | GPT-5.6 Terra63.4 |
| Knowledge | GPT-5.3 CodexNot measured | MarginNo overlap | GPT-5.6 Terra92.9 |
| Math | GPT-5.3 CodexNot measured | MarginNo overlap | GPT-5.6 Terra80.8 |
| Multimodal | GPT-5.3 CodexNot measured | MarginNo overlap | GPT-5.6 Terra80.7 |
Decisive benchmark drivers
The largest measured benchmark gaps in this matchup, with exact reported values.
More
- Source ↗
Terminal-Bench 2.0
AgenticA 77.3%B 87.4%Winner: GPT-5.6 TerraΔ 10.1Terminal-Bench 2.0: GPT-5.3 Codex scored 77.3%; GPT-5.6 Terra scored 87.4%. GPT-5.6 Terra wins this benchmark. - Source ↗
SWE-bench Pro
CodingA 56.8%B 63.4%Winner: GPT-5.6 TerraΔ 6.6SWE-bench Pro: GPT-5.3 Codex scored 56.8%; GPT-5.6 Terra scored 63.4%. GPT-5.6 Terra wins this benchmark.
Operational comparison
Runtime and commercial metrics are compared only when both models have a complete sourced value.
| Metric | GPT-5.3 Codex | GPT-5.6 Terra | Comparison |
|---|---|---|---|
| Input / output priceUSD per 1M tokens | GPT-5.3 Codex$1.75 input / $14 output | GPT-5.6 Terra$2.5 input / $15 output | GPT-5.3 Codex has the lower combined listed price. |
| Generation speedtokens per second | GPT-5.3 Codex79 tok/s | GPT-5.6 TerraNot available | A complete speed comparison is not available. |
| First-answer latencyseconds to first token | GPT-5.3 Codex88.26 s | GPT-5.6 TerraNot available | A complete latency comparison is not available. |
| Context windowmaximum listed tokens | GPT-5.3 Codex400K | GPT-5.6 Terra1M | GPT-5.6 Terra lists the larger context window. |
Benchmark Deep Dive
AgenticGPT-5.6 Terra wins20 benchmarks
| Benchmark | GPT-5.3 Codex | GPT-5.6 Terra | Result |
|---|---|---|---|
| Terminal-Bench 2.0Source | 77.3% | 87.4% | GPT-5.6 Terra leads |
| OSWorld-VerifiedSource | 64.7% | — | Not comparable |
| τ²-bench resultsSource | 86% | 86.3% | GPT-5.6 Terra leads |
| Gert LabsSource | 57.47% | — | Not comparable |
| JobBenchSource | 33.7% | — | Not comparable |
| BrowseCompSource | — | 87.5% | Not comparable |
| OSWorld 2.0Source | — | 50.2% | Not comparable |
| CyberGymSource | — | 81.8% | Not comparable |
| ExploitGymSource | — | 23.2% | Not comparable |
| ToolathlonSource | — | 53.1% | Not comparable |
| AA Agentic IndexSource | — | 47.4% | Not comparable |
| GDPval-AASource | — | 54.1% | Not comparable |
| GDPval-AASource | — | 1581 | Not comparable |
| AA Harvey LABSource | — | 85.2% | Not comparable |
| AA ITBenchSource | — | 51.0% | Not comparable |
| AA Tau3 BankingSource | — | 31.8% | Not comparable |
| AA AutomationBenchSource | — | 45.6% | Not comparable |
| terminalBenchHardSource | — | 57.6% | Not comparable |
| aaTerminalBench21Source | — | 88% | Not comparable |
| APEX-Agents-AASource | — | 38.9% | Not comparable |
CodingGPT-5.3 Codex wins10 benchmarks
| Benchmark | GPT-5.3 Codex | GPT-5.6 Terra | Result |
|---|---|---|---|
| SWE-bench VerifiedSource | 85% | — | Not comparable |
| SWE-bench ProSource | 56.8% | 63.4% | GPT-5.6 Terra leads |
| SWE-RebenchSource | 58.2% | — | Not comparable |
| Vibe Code BenchSource | 61.77% | — | Not comparable |
| AA-SciCodeSource | 53.2% | 53.9% | GPT-5.6 Terra leads |
| Terminal-Bench 2.0Source | — | 87.4% | Not comparable |
| deepSweSource | — | 69.6% | Not comparable |
| FrontierCode 1.1 ExtendedSource | — | 55.8% | Not comparable |
| cursorBench32Source | — | 64.9% | Not comparable |
| AA Coding IndexSource | — | 76.7% | Not comparable |
Reasoning3 benchmarks
Knowledge10 benchmarks
| Benchmark | GPT-5.3 Codex | GPT-5.6 Terra | Result |
|---|---|---|---|
| Artificial Analysis Intelligence IndexSource | 44.3% | 55.0% | GPT-5.6 Terra leads |
| AA-GPQA DiamondSource | 91.5% | 92.5% | GPT-5.6 Terra leads |
| AA-HLESource | 39.9% | 41.8% | GPT-5.6 Terra leads |
| AA-Omniscience IndexSource | 9.9% | -0.2% | GPT-5.3 Codex leads |
| AA-Omniscience AccuracySource | 51.8% | 45.9% | GPT-5.3 Codex leads |
| AA-Omniscience Hallucination RateSource | 86.9% | 85.2% | GPT-5.6 Terra leads |
| GPQASource | — | 92.9% | Not comparable |
| GPQA-DSource | — | 92.9% | Not comparable |
| HealthBench ProfessionalSource | — | 57.7% | Not comparable |
| HealthBench HardSource | — | 32.7% | Not comparable |
Math3 benchmarks
Multimodal4 benchmarks
Inst. Following1 benchmarks
| Benchmark | GPT-5.3 Codex | GPT-5.6 Terra | Result |
|---|---|---|---|
| AA-IFBenchSource | 75.4% | 71.2% | GPT-5.3 Codex leads |
Frequently Asked Questions (3)
Which is better, GPT-5.3 Codex or GPT-5.6 Terra?
GPT-5.6 Terra is ahead on BenchLM's BenchAlign leaderboard, 72.57 to 66.69. The biggest single separator in this matchup is Terminal-Bench 2.0, where the scores are 77.3% and 87.4%.
Which is better for coding, GPT-5.3 Codex or GPT-5.6 Terra?
GPT-5.3 Codex has the edge for coding in this comparison, averaging 67.2 versus 63.4. Inside this category, SWE-bench Pro is the benchmark that creates the most daylight between them.
Which is better for agentic tasks, GPT-5.3 Codex or GPT-5.6 Terra?
GPT-5.6 Terra has the edge for agentic tasks in this comparison, averaging 87.4 versus 71.4. Inside this category, Terminal-Bench 2.0 is the benchmark that creates the most daylight between them.
Related Comparisons
Explore More
Choose a model with this week’s evidence
Join 2,000+ readers for ranking moves, pricing changes, and the claims that still need proof.
One email each week. Unsubscribe anytime.