← Model list
Claude Opus 5
anthropic·anthropic/claude-opus-5·GA·Closed·claude-opus series·NEW
Strongest Claude Opus model for coding, agents, and professional work
At a glance
Official price$5.00 / $25.00 per 1M
Cache read $0.50 · Anthropic
Lowest paid$4.50 / $22.50
Requesty Gateway · 2.2× spread
Context1,000,000
Output limit128,000
Capabilities
✓ Reasoning✓ Tool use✓ Structured outputTemperature✓ Attachments
Modalities
TextImagePDF
Knowledge cutoff2026-05
Released / updated2026-07-24 / 2026-07-24
Quality & performanceArtificial Analysis · Intelligence Index v4.1 · rep. tier Adaptive Reasoning, Max Effort
Intelligence63.1
Value7
Coding78
Agentic59.2
Output speed53 tok/s
TTFT59.97 s
Cost per task$2.3369
Value formulaIQ 63.1 ÷ min blended $9.000 = 7
Reasoning tier → intelligence / speed (higher tier = stronger but slower)
Adaptive Reasoning, Low EffortIQ 52.5 · 49 tok/s
Adaptive Reasoning, Medium EffortIQ 58.6 · 50 tok/s
Adaptive Reasoning, High EffortIQ 61.5 · 51 tok/s
Adaptive Reasoning, Xhigh EffortIQ 62.5 · 54 tok/s
Adaptive Reasoning, Max EffortIQ 63.1 · 53 tok/s
Quality is independently evaluated by Artificial Analysis. Speed/latency are model-level medians.
Available at 28 providers27 with public prices
| Provider | Tier | Input | Output | Cache read | Cache write | Context | Output limit | Status |
|---|---|---|---|---|---|---|---|---|
| Requesty | Gateway | $4.50 | $22.50 | $0.45 | $5.63 | 1,000,000 | 128,000 | |
| NanoGPT | Gateway | $5.00 | $25.00 | $0.50 | $6.25 | 1,000,000 | 128,000 | |
| Vertex claude-opus-5@default | First-party | $5.00 | $25.00 | $0.50 | $6.25 | 1,000,000 | 128,000 | |
| OpenRouter | Gateway | $5.00 | $25.00 | $0.50 | $6.25 | 1,000,000 | 128,000 | |
| Vertex (Anthropic) claude-opus-5@default | Cloud | $5.00 | $25.00 | $0.50 | $6.25 | 1,000,000 | 128,000 | |
| CrossModel | Gateway | $5.00 | $25.00 | $0.50 | $6.25 | 1,000,000 | 128,000 | |
| AnthropicOfficial | First-party | $5.00 | $25.00 | $0.50 | $6.25 | 1,000,000 | 128,000 | |
| Amazon Bedrock anthropic.claude-opus-5 | Cloud | $5.00 | $25.00 | $0.50 | $6.25 | 1,000,000 | 128,000 | |
| Merge Gateway | Gateway | $5.00 | $25.00 | $0.50 | $6.25 | 1,000,000 | 128,000 | |
| Cloudflare AI Gateway | Cloud | $5.00 | $25.00 | $0.50 | $6.25 | 1,000,000 | 128,000 | |
Sorted by blended price (input×0.75 + output×0.25) asc. The official channel always shows regardless of rank. Whether a gateway's low price is actually usable can't be verified.
Your usage cost
1Requesty$1,539.00
2NanoGPT$1,710.00
3Vertex$1,710.00
4OpenRouter$1,710.00
5Vertex (Anthropic)$1,710.00
6CrossModel$1,710.00
7Anthropic · Official$1,710.00
Switch to Requesty to save $171.00/mo (10%). The gap is small, so staying on the official channel is fine.
Note: this is a gateway; verify availability and rate limits yourself.
Benchmark19 items
| Name | Conditions | Score | Metric | Source |
|---|---|---|---|---|
| SWE-Bench Verified | — | 96 | resolved | Source ↗ |
| SWE-Bench Pro | — | 79.2 | resolve rate | Source ↗ |
| SWE-Bench Multilingual | — | 89.5 | resolve rate | Source ↗ |
| SWE-Bench Multimodal | — | 59.4 | resolve rate | Source ↗ |
| DeepSWE | v1.1 | 68.8 | resolve rate | Source ↗ |
| FrontierCode | variant: medium effort · dataset: Main · v1.1 | 53.4 | mean@5 | Source ↗ |
| Frontier-Bench | harness: mini-SWE-agent · variant: max effort · vv0.1 | 43.3 | mean reward | Source ↗ |
| BrowseComp | variant: single agent | 90.8 | accuracy | Source ↗ |
| Humanity's Last Exam | variant: no tools | 56.3 | accuracy | Source ↗ |
| Humanity's Last Exam | variant: with tools | 64.7 | accuracy | Source ↗ |
| DeepSearchQA | variant: max effort | 95 | F1 | Source ↗ |
| OSWorld | v2.0 | 70.6 | success rate | Source ↗ |
| GDPval-AA | variant: max effort · vv2 | 1861 | Elo | Source ↗ |
| AA-Briefcase | variant: max effort | 1720 | Elo | Source ↗ |
| AutomationBench | variant: max effort | 26 | success rate | Source ↗ |
| ARC-AGI-1 | variant: max effort | 97.5 | accuracy | Source ↗ |
| ARC-AGI-2 | variant: max effort | 90.4 | accuracy | Source ↗ |
| ARC-AGI-3 | variant: high effort | 30.2 | RHAE | Source ↗ |
| HealthBench Professional | variant: max effort | 59.8 | score | Source ↗ |
The same benchmark scores very differently across harness / dataset, so the qualifying conditions must be shown together.
Reasoning control
effort = noneeffort = loweffort = mediumeffort = higheffort = maxbudget_tokens ≥ 1,024effort = xhighToggle (on / off)
3 / 28 providers expose no reasoning control (reasoning_options: []).
Experimental modes1 items
| Mode | Provider | Input | Output | Cache read | Cache write |
|---|---|---|---|---|---|
| fast | Anthropic | $10.00 | $50.00 | $1.00 | $12.50 |
Related models
Claude Opus 5 (Fast)same series$12.00 / $60.00Claude Opus 5 (EU)same series$5.50 / $27.50Claude Opus 5 (US)same series$5.00 / $25.00GLM-5.2cheaper alternative$1.40 / $4.40DeepSeek V4 Procheaper alternative$0.435 / $0.87DeepSeek V4 Flashcheaper alternative$0.14 / $0.28GPT-5.4cheaper alternative$2.50 / $15.00
Price historyone sample accumulated per data sync
Input listOutput listMin blended