Anthropic · Released Aug 5, 2025
Claude Opus 4.1
Claude Opus 4.1 is an updated version of Anthropic’s flagship model, offering improved performance in coding, reasoning, and agentic tasks. It achieves 74.5% on SWE-bench Verified and shows notable gains...
- Top 8%Lech Mazur Writing
- Top 24%GDPval
- Top 40%Cybench
- Top 40%SimpleBench
- Released
- Aug 5 2025
- Context window
- 200K 32K out
- Input / 1M tokens
- $15.00 from $13.00
- Output / 1M tokens
- $75.00 from $64.00
Where to run it
15 providers compared
Cheapest: Poe at $13.00 / $64.00 per 1M input / output tokens — 14% less than the priciest option.
| Provider | Input | Output | Cache read | Context | Max output | Uptime | vs. best |
|---|---|---|---|---|---|---|---|
| Poe | $13.00 | $64.00 | $1.30 | 197K | 32K | — | Cheapest |
| Amazon Bedrock | $15.00 | $75.00 | $1.50 | 200K | 32K | 100%24h | +17% |
| Helicone | $15.00 | $75.00 | $1.50 | 200K | 32K | — | +17% |
| NanoGPT | $15.00 | $75.00 | $1.50 | 200K | 32K | — | +17% |
| Kilo Gateway | $15.00 | $75.00 | $1.50 | 200K | 32K | — | +17% |
| Databricks | $15.00 | $75.00 | $1.50 | 200K | 32K | — | +17% |
| Pioneer | $15.00 | $75.00 | $1.50 | 200K | 32K | — | +17% |
| Requesty | $15.00 | $75.00 | $1.50 | 200K | 32K | — | +17% |
| FastRouter | $15.00 | $75.00 | $1.50 | 200K | 32K | — | +17% |
| DigitalOcean | $15.00 | $75.00 | $1.50 | 200K | 8K | — | +17% |
| ZenMux | $15.00 | $75.00 | $1.50 | 200K | 32K | — | +17% |
| Azure | $15.00 | $75.00 | $1.50 | 200K | 32K | — | +17% |
| Neon | $15.00 | $75.00 | $1.50 | 200K | 32K | — | +17% |
| Azure Cognitive Services | $15.00 | $75.00 | $1.50 | 200K | 32K | — | +17% |
| OpenCode | $15.00 | $75.00 | $1.50 | 200K | 32K | — | +17% |
Value
Price vs. performance
Claude Opus 4.1 Other Anthropic Other labs Value frontier · GPQA Diamond vs. input price per 1M tokens (log scale) · Drag to pan, pinch or Ctrl + scroll to zoom, double-click to zoom in.
Benchmarks
16 benchmark scores
| Benchmark | Score | Rank | Leader |
|---|---|---|---|
| SWE-bench VerifiedSoftware engineering · Real GitHub issues fixed in Python repos | 73.4% | #19of 32 | Claude Opus 4.783.5% |
| Terminal-BenchSoftware engineering · Tasks completed with the programs in a real terminal (best agent) | 38.0% | #25of 40 | GPT-5.584.7% |
| METR time horizonSoftware engineering · Length of software task solved more often than not | 1.9 h | #11of 23 | Claude Opus 4.612 h |
| GDPvalAgents & computer use · Real-world occupational tasks, win rate vs experts | 43.6% | #3of 13 | GPT-5.249.7% |
| Lech Mazur WritingWriting & content · Short stories weaving 10 required elements, judged 0–10 | 8.47/10 | #2of 27 | GPT-58.60/10 |
| DeepResearch BenchWriting & content · Find and synthesize information from the web | 48.3% | #9of 21 | Claude Opus 4.655.3% |
| WebDev ArenaDesign & visual · People vote on which model builds the better website (Elo) | 1389 | #66of 90 | Claude Opus 5.51820 |
| VPCTDesign & visual · Read diagrams of balls rolling down ramps | 35.0% | #16of 18 | GPT-5.284.0% |
| CybenchCybersecurity · Find and exploit vulnerabilities in sandboxed challenges | 42.0% | #4of 10 | Claude Opus 4.693.0% |
| GPQA DiamondReasoning & knowledge · PhD-level biology, chemistry and physics questions | 77.3% | #76of 122 | Gemini 4 Argon96.4% |
| Humanity's Last ExamReasoning & knowledge · Expert questions across 100+ subjects | 11.5% | #21of 32 | GPT-6 Astra54.8% |
| SimpleBenchReasoning & knowledge · Common-sense trick questions about space, time and people | 60.0% | #23of 58 | Claude Fable 581.9% |
| LMCAReasoning & knowledge · Judge the quality of conceptual arguments like experts | 37.1% | #62of 123 | Claude Sonnet 5.568.3% |
| FrontierMathMaths · Expert-written problems, undergrad to early research (tiers 1–3) | 12.6% | #76of 87 | GPT-6.1 Sol93.7% |
| FrontierMath Tier 4Maths · Exceptionally hard research-level problems | 2.4% | #58of 64 | GPT-6.1 Sol100.0% |
| OTIS Mock AIMEMaths · Competition-style problems harder than MATH Level 5 | 68.9% | #83of 120 | Gemini 4 Argon100.0% |
Specs
Details
- Model ID
anthropic/claude-opus-4.1- Developer
- Anthropic
- Release date
- Aug 5, 2025 ·
- Context window
- 200,000 tokens
- Max output
- 32,000 tokens
- API price
- $15.00 / $75.00 per 1M input / output tokens
- Providers
- 15
- Input types
- image, text, file
- Reasoning
- Yes
- Tool calling
- Yes
- Open weights
- No
- Free tier
- No
Release date and prices from OpenRouter's catalog; providers from OpenRouter and models.dev. Buying directly from a provider can cost a little differently. Speed, latency and uptime are measured on OpenRouter only; uptime is the share of OpenRouter requests each provider served over the last day. Benchmark scores are the best result per benchmark from Epoch AI, licensed CC BY 4.0. Rank counts every model tracked here with a score.