// Anthropic
Claude Opus 4.5
A Anthropic model. 1 prices across 1 channels side by side, each with its source and check date – plus specs, licence, speed and benchmarks.
- Official in/out per 1M
- $5 / $25
- Lowest (official)
- $5 / $25
- Context
- 200K
- AA index
- 23.7
Prices checked 2026-10-10Official sources linked on every rowNo estimatesCheapest LLM APIs →
01Specs at a glance
Official values first; gaps filled from OpenRouter’s API where the vendor page is silent.
| Vendor | Anthropic |
|---|---|
| Context window | 200K (200,000 tokens) |
| Max output | 64K |
| Input | File/PDF · Image · Text |
| Output | Text |
| Tool calling / JSON | Yes / Yes |
| Weights & licence | Closed · Proprietary (API only) |
| Released | Nov 24, 2025 (OpenRouter listing date) |
| Model IDs | claude-opus-4-5 |
02Prices on 1 channels
Per 1M tokens, in the currency each source publishes (1 USD = 6.7 CNY for the blended column). Sorted: official first, then cheapest.
| Channel | Source | Input | Output | Cache hit | Blended $ | Context | Notes |
|---|---|---|---|---|---|---|---|
| Official Vendor API, billed in USD | Official 10-10 | $5 | $25 | $0.5 | 10 | 200K | Batch −50%Batch -50%; cache write 5m=1.25x input, 1h=2x input |
“Official” = vendor pricing page; “Platform” = the platform’s own pricing page; “OpenRouter” = price listed on OpenRouter, not checked on the platform’s site. Click a source tag to open the original page.
03Where you can use it
Derived only from the price sources above. Country restrictions are set by each vendor – check their terms before you sign up.
- Vendor API, billed in USD
Official
04Performance
All results →Speed and AA index from Artificial Analysis public leaderboards (not our own tests); benchmark scores from our benchmark library, best result per benchmark.
- AA indexClaude Opus 4.5 (Non-reasoning)
- 23.7
- Output speed (median)
- 48 tok/s
- Time to first token
- 0.96s
- AA index per $ (blended)
- 2.37
| Benchmark | Best score | Source | Date |
|---|---|---|---|
| LMArena Text | 1474 | official | 2026-10-08 |
| LMArena WebDev | 1494 | official | 2026-10-08 |
| GPQA Diamond | 86.0% | independent | 2025-11-24 |
| SWE-bench Verified | 79.2% | official | 2025-12-05 |
| AIME (OTIS Mock AIME 2024–2025) | 86.1% | independent | 2025-11-24 |
| Humanity's Last Exam | 25.2% | compiled | 2025-11-24 |
| OSWorld (Verified) | 66.3% | vendor | 2025-11-25 |
| τ²-bench | 92.3% | official | 2026-05-05 |
| Terminal-Bench 2.0 | 63.1% | official | 2025-11-24 |
| BFCL v4 (function calling) | 77.5% | official | 2026-10-17 |
05Price changes we recorded
All price changes →From our daily re-checks of the sources above.
No price change recorded since we started tracking (2026-10-10).
06More from Anthropic
FAQ
How much does the Claude Opus 4.5 API cost?
On the official pricing page: $5 / $25 per 1M input / output tokens, cached input $0.5 (checked 2026-10-10).
What is the context window of Claude Opus 4.5?
200K tokens of context, up to 64K output tokens. Platforms may cap it lower – see the table.
Is Claude Opus 4.5 open source?
No. It is only available through APIs and the vendor’s apps.