Claude Opus 4.7 vs Claude Opus 4.8
| Spec | Claude Opus 4.7 | Claude Opus 4.8 |
|---|---|---|
| Respan Index | 50.0 (#43) | 55.7 (#27) |
| Blended price per 1M | $10 | $10 |
| Context window | 1M | 1M |
| Input / output per 1M | $5 / $25 | $5 / $25 |
| Max output | 128K | 128K |
| Input | text, vision | text, vision |
| Released | 2026-04-16 | 2026-05-28 |
Respan Index by category
Each category is scored 0 to 100, and how far ahead the winner is. Leads under a point are close to even.
| Category | Claude Opus 4.7 | Claude Opus 4.8 |
|---|---|---|
| Coding | 32.2 Claude Opus 4.7 | 39.4 Claude Opus 4.8 |
| Agents | 61.1 Claude Opus 4.7 | 64.4 Claude Opus 4.8 |
| Reasoning | 53.6 Claude Opus 4.7 | 64.8 Claude Opus 4.8 |
| Knowledge & language | 49.9 Claude Opus 4.7 | 52.5 Claude Opus 4.8 |
| Human preference | 55.4 Claude Opus 4.7 | 55.2 Claude Opus 4.8 |
Benchmarks both report
Claude Opus 4.8 leads on 26 of the 37 benchmarks both report. Higher is better. Each score links to where it was published.
| Benchmark | Claude Opus 4.7 | Claude Opus 4.8 |
|---|---|---|
| LMArena Elo | 1501 | 1482 |
| LMArena Vision (LMArena) | 1299 | 1286 |
| LMArena WebDev (LMArena) | 1557 | 1555 |
| MirrorCode (Epoch AI) | 43.3% | 65.6% |
| Chess Puzzles (Epoch AI) | 41% | 52% |
| APEX-Agents (Mercor) | 49.2% | 48.9% |
| EBR-bench (Epoch AI) | 19% | 28.6% |
| FrontierMath Tier 4 v2 (Epoch AI) | 31.7% | 56.1% |
| FrontierMath Tiers 1-3 v2 (Epoch AI) | 70.2% | 80% |
| Furniture Assembly (Epoch AI) | 33.3% | 42.5% |
| GPQA Diamond | 94.2% | 93.6% |
| LiveBench Agentic Coding (LiveBench) | 50.7% | 50.5% |
| LiveBench Coding (LiveBench) | 82.1% | 81.8% |
| LiveBench Data Analysis (LiveBench) | 78.3% | 66% |
| LiveBench Instruction Following (LiveBench) | 66.7% | 72% |
| LiveBench Language (LiveBench) | 77.9% | 79.7% |
| LiveBench Mathematics (LiveBench) | 92.9% | 94.3% |
| LiveBench Reasoning (LiveBench) | 87.2% | 89.2% |
| Mystery Game Puzzles (Epoch AI) | 28% | 36% |
| OTIS Mock AIME 2024-2025 (Epoch AI) | 97.8% | 98.3% |
| SAGE (Vals AI) | 56.1% | 54.8% |
| SimpleQA Verified | 51.7% | 53% |
| Vending-Bench 2 (Andon Labs) | 10936.76 | 5787.43 |
| WeirdML (Håvard Tveit Ihle) | 76.4% | 82.9% |
| AIME 2026 | 95.8% | 100% |
| BrowseComp | 79.8% | 84.3% |
| Humanity's Last Exam (no tools) | 46.9% | 49.8% |
| Humanity's Last Exam (with tools) | 54.7% | 57.9% |
| OSWorld-Verified | 82.8% | 83.4% |
| SWE-bench Multilingual | 80.5% | 84.4% |
| SWE-bench Multimodal | 34.5% | 38.4% |
| SWE-Bench Pro | 64.3% | 69.2% |
| SWE-bench Verified | 87.6% | 88.6% |
| Terminal-Bench 2.1 | 66.1% | 74.6% |
| tau2-bench Banking Knowledge (Sierra) | 40.2% | 39.7% |
| GSO Opt@1 (GSO) | 42.2% | 47.1% |
| SimpleBench (SimpleBench) | 61.7% | 64.8% |
FAQ
Is Claude Opus 4.7 better than Claude Opus 4.8?
Claude Opus 4.8 scores 55.7 on the Respan Index (#27), ahead of Claude Opus 4.7 at 50.0 (#43). Claude Opus 4.8 leads on 26 of the 37 benchmarks both report.
Which is cheaper, Claude Opus 4.7 or Claude Opus 4.8?
Claude Opus 4.7 lists at $5 input and $25 output per 1M tokens; Claude Opus 4.8 at $5 and $25. Both cost about the same per token at list price.
Which has the larger context window?
Both accept 1M tokens of context.
Compare other models
- vsClaude Opus 4.7 vs Claude Opus 4.6
- vsClaude Opus 5 vs Claude Opus 4.8
- vsClaude Opus 4.8 vs NVIDIA Nemotron 3 Ultra 550B-A55B NVFP4
- vsClaude Opus 4.7 vs GPT-5.4 Pro
- vsClaude Opus 4.7 vs GPT-5.4
- vsClaude Opus 4.7 vs MiMo-V2.6-Flash
- vsClaude Opus 4.8 vs Muse Spark 1.3
- vsClaude Opus 4.8 vs Step 5 Preview
- vsClaude Opus 4.8 vs MiMo-V2.6-Pro