Claude Opus 5 vs Claude Sonnet 5.5
Different efforts: medium · high · Compare all at high
AutomationBench-AA AutomationBench-AA: Agents orchestrate SaaS workflows through APIs; objectives receive credit only when the task has no guardrail violation
Opus 554.3
Sonnet 5.559.4
GDPval-AA (work tasks) 2.1 GDPval-AA (work tasks) 2.1: Real work tasks from many professions (reports, spreadsheets, analyses), graded against expert work
Opus 549.6
Sonnet 5.552.5
AA Analyst Agent AA Analyst Agent: Data-analyst tasks: the model explores data with tools and reports findings
Opus 5 · max53.8
Sonnet 5.5 · max57.5
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
Opus 534.3
Sonnet 5.543.9
Terminal-Bench Science 0.1 Terminal-Bench Science 0.1: Agents complete expert-authored research workflows across five scientific domains in a terminal
Opus 5 · max28.6
Sonnet 5.534.3
AA-Briefcase 1.1 · Elo AA-Briefcase 1.1 · Elo: Agents produce professional deliverables; blind pairwise comparisons produce the source Elo rating
Opus 51,434.9
Sonnet 5.51,635.8
AA-LCR (long context) 1.1 AA-LCR (long context) 1.1: Questions that need reading and connecting facts across very long documents (around 100K tokens)
Opus 582
Sonnet 5.578
DeepSWE 1.1 DeepSWE 1.1: Agents change repository code to resolve software engineering tasks under the stated DeepSWE edition
Opus 568.9
Sonnet 5.5 · max70.5
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
Opus 5 · xhigh99.4
Sonnet 5.5 · xhigh95.8
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
Opus 5 · xhigh98
Sonnet 5.5 · xhigh70.6
Furniture Assembly Furniture Assembly: Models inspect furniture assembly photos, identify incorrect builds and locate the mistaken step
Opus 5 · max60.8
Sonnet 5.5 · max75
τ³-bench · banking · 1.0.1 τ-bench official · 3-banking-1.0.1: Customer-service agents converse with simulated users, retrieve domain knowledge and change state through tools
Opus 5 · max48.7
Sonnet 5.5—
τ³-Banking · AA τ³-Banking · AA: Agents retrieve banking knowledge and execute tools to resolve customer-support workflows
Opus 538.6
Sonnet 5.5—
EnterpriseOps-Gym-AA EnterpriseOps-Gym-AA: Agents complete stateful enterprise workflows in eight domains; final database state determines success
Opus 5 · max47.5
Sonnet 5.5—
FrontierCode 1.1 FrontierCode 1.1: Agents implement changes in real repositories, graded against maintainer-defined requirements
Opus 5—
Sonnet 5.544.3
Harvey LAB-AA 1.1 Harvey LAB-AA 1.1: Legal agents work from case documents; every rubric criterion must pass and material hallucinations receive no credit
Opus 5—
Sonnet 5.5 · max2.8
EBR Bench EBR Bench: Agents learn to play Earthborne Rangers across repeated playthroughs
Opus 5 · max45.7
Sonnet 5.5—
SpireClimb SpireClimb: Agents play Slay the Spire under the named game and difficulty setting
Opus 5—
Sonnet 5.5 · max1.3
SpireClimb StS1 A20 Heart SpireClimb StS1 A20 Heart: Agents play Slay the Spire under the named game and difficulty setting
Opus 5—
Sonnet 5.5 · max0
SpireClimb StS2 A10 SpireClimb StS2 A10: Agents play Slay the Spire under the named game and difficulty setting
Opus 5—
Sonnet 5.5 · max2.7
| Benchmark | Opus 5 | Sonnet 5.5 |
|---|---|---|
| AutomationBench-AA | 54.3 | 59.4 |
| GDPval-AA (work tasks) 2.1 | 49.6 | 52.5 |
| AA Analyst Agent | 53.8 · max | 57.5 · max |
| Terminal-Bench 4.0 | 34.3 | 43.9 |
| Terminal-Bench Science 0.1 | 28.6 · max | 34.3 |
| AA-Briefcase 1.1 · Elo | 1,434.9 | 1,635.8 |
| AA-LCR (long context) 1.1 | 82 | 78 |
| DeepSWE 1.1 | 68.9 | 70.5 · max |
| SWE-bench Pro (v2 full) | 99.4 · xhigh | 95.8 · xhigh |
| SWE-bench Pro (v2 hard) | 98 · xhigh | 70.6 · xhigh |
| Furniture Assembly | 60.8 · max | 75 · max |
| τ³-bench · banking · 1.0.1 | 48.7 · max | — |
| τ³-Banking · AA | 38.6 | — |
| EnterpriseOps-Gym-AA | 47.5 · max | — |
| FrontierCode 1.1 | — | 44.3 |
| Harvey LAB-AA 1.1 | — | 2.8 · max |
| EBR Bench | 45.7 · max | — |
| SpireClimb | — | 1.3 · max |
| SpireClimb StS1 A20 Heart | — | 0 · max |
| SpireClimb StS2 A10 | — | 2.7 · max |
Specs
| Spec | Opus 5 | Sonnet 5.5 |
|---|---|---|
| Input, per 1M | $5 | $2 |
| Output, per 1M | $25 | $10 |
| Cost per task? | $2.19 | $0.88 |
| Speed, tok/s | 55 | 106 |
| Context | 1M | 1M |
| Weights | Closed | Closed |
| Released | Jul 2026 | Sep 2026 |
| See images | ✓ | ✓ |
| Read PDFs | Partial | ✓ |
| Hear audio | — | — |
| Watch video | — | — |
| Call tools | ✓ | ✓ |
| Force a tool call | Partial | — |
| Return strict JSON | ✓ | ✓ |
| Search the web | ? | ✓ |
| Run code | ? | ✓ |
| Use a computer | ? | ✓ |
What breaks when you switch
Opus 5 → Sonnet 5.5
No forced tool use›
Forced tool_choice any or tool returns HTTP 400; remove the forced choice.
Replace disabled thinking›
Disabled thinking returns HTTP 400; use between_tools for the lowest thinking setting.
No manual thinking budgets›
Manual thinking budgets and extra between_tools fields return HTTP 400; remove manual budgets and send between_tools without display, budget_tokens or block_binding.
Remove sampling parameters›
Non-default temperature, top_p or top_k returns HTTP 400; remove these parameters.
No assistant prefill›
A prefilled final assistant turn returns HTTP 400; remove that prefill.
Old computer tool rejected · Claude API / Google Cloud›
The computer_20251124 tool returns HTTP 400 on the Claude API and Google Cloud; use computer_toolset_20260801 there.
Sonnet 5.5 → Opus 5
Nothing known.