Claude Sonnet 5 vs GPT-5.6 Terra
GPT-5.6 Terra leads Claude Sonnet 5 by 6.4 points on the Shogo Agent Index, but their ranges overlap, so treat it as close. Claude Sonnet 5 wins 5 and GPT-5.6 Terra wins 14 of 19 shared benchmarks. Claude Sonnet 5 is cheaper for agent workloads.
Add more models to this comparison →Pillars
| Pillar | Claude Sonnet 5 | GPT-5.6 Terra |
|---|---|---|
| Agentic coding | 57 | 66 |
| Long-horizon autonomy | 65 | 69 |
| Knowledge-work agents | 43 | 47 |
| Computer use | 40 | 48 |
| Security & context learning | 51 | 56 |
Shared benchmarks
| Benchmark | Claude Sonnet 5 | GPT-5.6 Terra |
|---|---|---|
| APEX-Agentsindex | 54.5% | 58.2% |
| DeepSWEindex | 53.8% | 69.6% |
| Vending-Bench 2index | $6,378 | $7,343 |
| CursorBenchindex | 34.1% | 41.3% |
| Chess Puzzles | 45.0% | 47.0% |
| CritPt | 16.9% | 30.0% |
| DTBench | 92.5% | 93.3% |
| FrontierCode | 42.7% | 41.3% |
| FrontierMath Tier 4 | 29.3% | 70.7% |
| FrontierMath Tiers 1-3 | 65.6% | 86.0% |
| GPQA Diamond | 90.5% | 93.3% |
| LMCA | 50.0% | 54.9% |
| MirrorCode | 63.3% | 41.1% |
| OTIS Mock AIME | 94.7% | 99.7% |
| ProofBench | 77.0% | 74.0% |
| SciCode | 54.3% | 55.0% |
| SimpleBench | 60.6% | 48.9% |
| SimpleQA Verified | 33.7% | 43.2% |
| WebDev Arena | 1537 | 1521 |
Specs and price
| Claude Sonnet 5 | GPT-5.6 Terra | |
|---|---|---|
| Agent Index | 52.4 (#32) | 58.8 (#16) |
| Agent $/M | $0.92 | $1.02 |
| Input / output $/M | $2.00 / $10.00 | $2.00 / $12.00 |
| Context | 1M | 1.1M |
| Released | Jun 30, 2026 | Jul 9, 2026 |
| Weights | Closed | Closed |
| API hosts | 5 | 4 |
Don’t pick one. Route.
Shogo can run Claude Sonnet 5 and GPT-5.6 Terra side by side and send each step of an agent to the one that’s cheaper for it.
Try both in Shogo