Skip to content

Anthropic · Model profile

Claude Opus 4.5

Intelligence, Performance & Price Analysis

Source: Canonical slug: claude-opus-4-5 · Canonical model registry at build time

Routing verdict

Frontier research / complex reasoning

Reserve this model for the hardest requests; simpler or repetitive work should stay on cheaper routes.

Claude Opus 4.5 summary metrics

Intelligence
35 scoreClaude Opus 4.5 (Non-reasoning) scores 35 (estimated) on the Artificial Analysis Intelligence Index, placing it well above average among other non-reasoning models in a similar price tier (median: 17).
Speed
51.3 output tokens/secClaude Opus 4.5 (Non-reasoning) generates output at 51.3 tokens per second (based on Anthropic's API), which is below average compared to other non-reasoning models in a similar price tier (median: 64.2 t/s).
Latency
1.51s TTFTClaude Opus 4.5 (Non-reasoning) has a time to first token (TTFT) of 1.51s (based on Anthropic's API), which is better than average compared to other non-reasoning models in a similar price tier (median: 1.56s).
Input Price
$5.00 / 1M tokensClaude Opus 4.5 (Non-reasoning) costs $5.00 per 1M input tokens (at the higher end, median: $1.88) and $25.00 per 1M output tokens (at the higher end, median: $7.75), based on Anthropic's API.
Output Price
$25.00 / 1M tokensClaude Opus 4.5 (Non-reasoning) costs $5.00 per 1M input tokens (at the higher end, median: $1.88) and $25.00 per 1M output tokens (at the higher end, median: $7.75), based on Anthropic's API.

Assessment

Routing verdict and task fit

Claude Opus 4.5 (Non-reasoning) scores 35 (estimated) on the Artificial Analysis Intelligence Index, placing it well above average among other non-reasoning models in a similar price tier (median: 17). Claude Opus 4.5 (Non-reasoning) generates output at 51.3 tokens per second (based on Anthropic's API), which is below average compared to other non-reasoning models in a similar price tier (median: 64.2 t/s). Claude Opus 4.5 (Non-reasoning) costs $5.00 per 1M input tokens (at the higher end, median: $1.88) and $25.00 per 1M output tokens (at the higher end, median: $7.75), based on Anthropic's API.
Task fit assessment
WorkloadRatingNotes
Complex reasoning & agentic workflowsOptimalIntelligence score 35 places this model among top performers. Suitable for multi-step analysis and agentic loops.
High-volume chat & customer-facingOptimalOutput speed 51.3 tokens/sec is adequate for chat.
Latency-sensitive applicationsOptimalTTFT 1.51s — adequate latency for most interactive use cases.
Cost-sensitive pipelinesViableOutput pricing at $25.00 is premium. Route high-volume simple tasks to cheaper alternatives.

Evidence

Evidence charts

14 published comparison charts for this profile. Claude Opus 4.5 is marked with a diamond and bold label; each chart has a table view and lists its source and retrieval date.

AA-Omniscience Index

AA-Omniscience Index (higher is better) measures knowledge reliability and hallucination. It rewards correct answers, penalizes hallucinations, and has no penalty for refusing to answer. Scores range from -100 to 100, where 0 means as many correct as incorrect answers, and negative scores mean more incorrect than correct. · Evaluation results measured independently by Artificial Analysis

19 models in the published order. Claude Opus 4.5: -4, row 11 of 19 in the published order.
  1. Claude Fable 5 (with fallback)40.1
  2. Gemini 3.5 Flash22.7
  3. GPT-5.5 (xhigh)20.1
  4. Grok 4.3 (high)18.3
  5. Kimi K2.66.4
  6. Muse Spark4.1
  7. GLM-5.2 (max)4
  8. MiniMax-M31.4
  9. Nemotron 3 Ultra-0.833
  10. Claude Sonnet 5 (Non-reasoning)-1.2
  11. Claude Opus 4.5 (this model)-4
  12. GPT-5.5 (Non-reasoning)-5
  13. DeepSeek V4 Pro (max)-10
  14. Claude 3.5 Haiku-23.2
  15. Grok 4.3 (Non-reasoning)-32.3
  16. DeepSeek V3 0324-37.6
  17. Nova 2.0 Pro Preview-48.1
  18. gpt-oss-120b (high)-50
  19. Qwen3.6 27B-52.8

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Chart interpretation
Each bar is one model measured the same way, so this model can be read against the others. The chart lists its source and retrieval date.

Artificial Analysis Intelligence Index by Open Weights / Proprietary

Artificial Analysis Intelligence Index v4.1 incorporates 9 evaluations: GDPval-AA v2, 𝜏³-Banking, Terminal-Bench v2.1, SciCode, Humanity's Last Exam, GPQA Diamond, CritPt, AA-Omniscience, AA-LCR · Evaluation results measured independently by Artificial Analysis

19 models in the published order. Claude Opus 4.5: 34.7, row 13 of 19 in the published order.
  1. Claude Fable 5 (with fallback)59.9
  2. GPT-5.5 (xhigh)54.8
  3. GLM-5.2 (max)51.1
  4. Gemini 3.5 Flash50.2
  5. MiniMax-M344.4
  6. DeepSeek V4 Pro (max)44.3
  7. Kimi K2.644.2
  8. Muse Spark43.1
  9. Claude Sonnet 5 (Non-reasoning)41.7
  10. Nemotron 3 Ultra37.8
  11. Grok 4.3 (high)37.6
  12. GPT-5.5 (Non-reasoning)35.4
  13. Claude Opus 4.5 (this model)34.7
  14. Qwen3.6 27B30.5
  15. Grok 4.3 (Non-reasoning)24.8
  16. gpt-oss-120b (high)23.8
  17. DeepSeek V3 032415.4
  18. Nova 2.0 Pro Preview14.4
  19. Claude 3.5 Haiku12.3

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Benchmark interpretation
Higher values indicate stronger overall capability. Bars show how models compare on composite intelligence metrics.

Artificial Analysis Intelligence Index

Artificial Analysis Intelligence Index v4.1 incorporates 9 evaluations: GDPval-AA v2, 𝜏³-Banking, Terminal-Bench v2.1, SciCode, Humanity's Last Exam, GPQA Diamond, CritPt, AA-Omniscience, AA-LCR · Evaluation results measured independently by Artificial Analysis

19 models in the published order. Claude Opus 4.5: 34.7, row 13 of 19 in the published order.
  1. Claude Fable 5 (with fallback)59.9
  2. GPT-5.5 (xhigh)54.8
  3. GLM-5.2 (max)51.1
  4. Gemini 3.5 Flash50.2
  5. MiniMax-M344.4
  6. DeepSeek V4 Pro (max)44.3
  7. Kimi K2.644.2
  8. Muse Spark43.1
  9. Claude Sonnet 5 (Non-reasoning)41.7
  10. Nemotron 3 Ultra37.8
  11. Grok 4.3 (high)37.6
  12. GPT-5.5 (Non-reasoning)35.4
  13. Claude Opus 4.5 (this model)34.7
  14. Qwen3.6 27B30.5
  15. Grok 4.3 (Non-reasoning)24.8
  16. gpt-oss-120b (high)23.8
  17. DeepSeek V3 032415.4
  18. Nova 2.0 Pro Preview14.4
  19. Claude 3.5 Haiku12.3

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Benchmark interpretation
Higher values indicate stronger overall capability. Bars show how models compare on composite intelligence metrics.

Intelligence

Artificial Analysis Intelligence Index · Higher is better · Evaluation results measured independently by Artificial Analysis

12 models in the published order. Claude Opus 4.5: 34.7, row 11 of 12 in the published order.
  1. Claude Fable 5 (with fallback)59.9
  2. GPT-5.5 (xhigh)54.8
  3. GLM-5.2 (max)51.1
  4. Gemini 3.5 Flash50.2
  5. MiniMax-M344.4
  6. DeepSeek V4 Pro (max)44.3
  7. Kimi K2.644.2
  8. Muse Spark43.1
  9. Nemotron 3 Ultra37.8
  10. Grok 4.3 (high)37.6
  11. Claude Opus 4.5 (this model)34.7
  12. gpt-oss-120b (high)23.8

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Benchmark interpretation
Higher values indicate stronger overall capability. Bars show how models compare on composite intelligence metrics.

Output Speed

Output tokens per second · Higher is better · Evaluation results measured independently by Artificial Analysis

15 models in the published order. Claude Opus 4.5: 51 t/s, row 15 of 15 in the published order.
  1. gpt-oss-120b (high)316 t/s
  2. Nemotron 3 Ultra248 t/s
  3. GLM-5.2 (max)216 t/s
  4. Gemini 3.5 Flash192 t/s
  5. Grok 4.3 (high)164 t/s
  6. Nova 2.0 Pro Preview141 t/s
  7. Grok 4.3 (Non-reasoning)133 t/s
  8. MiniMax-M395 t/s
  9. GPT-5.5 (xhigh)88 t/s
  10. Kimi K2.676 t/s
  11. GPT-5.5 (Non-reasoning)75 t/s
  12. DeepSeek V4 Pro (max)72 t/s
  13. Claude Fable 5 (with fallback)71 t/s
  14. Qwen3.6 27B60 t/s
  15. Claude Opus 4.5 (this model)51 t/s

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Output speed context
Speed should be read together with latency and cost; fast output does not always mean better user-perceived performance.

Speed

Output tokens per second · Higher is better · Evaluation results measured independently by Artificial Analysis

11 models in the published order. Claude Opus 4.5: 51 t/s, row 11 of 11 in the published order.
  1. gpt-oss-120b (high)316 t/s
  2. Nemotron 3 Ultra248 t/s
  3. GLM-5.2 (max)216 t/s
  4. Gemini 3.5 Flash192 t/s
  5. Grok 4.3 (high)164 t/s
  6. MiniMax-M395 t/s
  7. GPT-5.5 (xhigh)88 t/s
  8. Kimi K2.676 t/s
  9. DeepSeek V4 Pro (max)72 t/s
  10. Claude Fable 5 (with fallback)71 t/s
  11. Claude Opus 4.5 (this model)51 t/s

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Output speed context
Speed should be read together with latency and cost; fast output does not always mean better user-perceived performance.

End-to-End Response Time

Seconds to output 500 tokens, including reasoning model 'thinking' time · Lower is better · Evaluation results measured independently by Artificial Analysis

15 models in the published order. Claude Opus 4.5: 9.74 s, row 5 of 15 in the published order.
  1. Grok 4.3 (Non-reasoning)3.76 s
  2. Nova 2.0 Pro Preview3.53 s
  3. GPT-5.5 (Non-reasoning)6.65 s
  4. gpt-oss-120b (high)1.58 s
  5. Claude Opus 4.5 (this model)9.74 s
  6. Qwen3.6 27B8.39 s
  7. Nemotron 3 Ultra2.02 s
  8. GLM-5.2 (max)2.31 s
  9. Grok 4.3 (high)3.05 s
  10. MiniMax-M35.25 s
  11. Gemini 3.5 Flash2.6 s
  12. Kimi K2.66.56 s
  13. DeepSeek V4 Pro (max)6.95 s
  14. GPT-5.5 (xhigh)5.68 s
  15. Claude Fable 5 (with fallback)7.07 s

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Chart interpretation
Each bar is one model measured the same way, so this model can be read against the others. The chart lists its source and retrieval date.

Latency: Time To First Answer Token

Seconds to first answer token received · Accounts for reasoning model 'thinking' time · Evaluation results measured independently by Artificial Analysis

15 models in the published order. Claude Opus 4.5: 0 s, row 4 of 15 in the published order.
  1. Grok 4.3 (Non-reasoning)0 s
  2. GPT-5.5 (Non-reasoning)0 s
  3. Nova 2.0 Pro Preview0 s
  4. Claude Opus 4.5 (this model)0 s
  5. Qwen3.6 27B0 s
  6. gpt-oss-120b (high)6.33 s
  7. Nemotron 3 Ultra9.17 s
  8. GLM-5.2 (max)9.26 s
  9. Grok 4.3 (high)0 s
  10. MiniMax-M320.98 s
  11. Gemini 3.5 Flash0 s
  12. Kimi K2.658.43 s
  13. DeepSeek V4 Pro (max)60.79 s
  14. GPT-5.5 (xhigh)0 s
  15. Claude Fable 5 (with fallback)0 s

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Latency considerations
Latency includes thinking / reasoning time when applicable. Lower is better for interactive use cases.

Context Window

Context window: tokens limit · Higher is better · Evaluation results measured independently by Artificial Analysis

19 models in the published order. Claude Opus 4.5: 200k, row 16 of 19 in the published order.
  1. Gemini 3.5 Flash1M
  2. Claude Fable 5 (with fallback)1M
  3. GLM-5.2 (max)1M
  4. DeepSeek V4 Pro (max)1M
  5. Grok 4.3 (high)1M
  6. MiniMax-M31M
  7. Claude Sonnet 5 (Non-reasoning)1M
  8. Grok 4.3 (Non-reasoning)1M
  9. GPT-5.5 (xhigh)922k
  10. GPT-5.5 (Non-reasoning)922k
  11. Nemotron 3 Ultra262.1k
  12. Muse Spark262.1k
  13. Qwen3.6 27B262.1k
  14. Kimi K2.6256k
  15. Nova 2.0 Pro Preview256k
  16. Claude Opus 4.5 (this model)200k
  17. Claude 3.5 Haiku200k
  18. gpt-oss-120b (high)131.1k
  19. DeepSeek V3 0324128k

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Context window sizing
Larger context windows enable complex multi-turn tasks but may increase latency and cost for long inputs.

Cost per Intelligence Index Task

Weighted average cost (USD) per Artificial Analysis Intelligence Index task, segmented by token type. Lower is better · Evaluation results measured independently by Artificial Analysis

15 models in the published order. Claude Opus 4.5 is not marked in this published set of 15 models.
  1. DeepSeek V4 Pro (max)0.0063
  2. gpt-oss-120b (high)0.00454
  3. MiniMax-M30.0139
  4. Grok 4.3 (high)0.0101
  5. GPT-5.5 (Non-reasoning)0.0476
  6. DeepSeek V3 03240.00334
  7. Nemotron 3 Ultra0.0209
  8. Nova 2.0 Pro Preview0.0363
  9. Grok 4.3 (Non-reasoning)0.023
  10. Kimi K2.60.0216
  11. Qwen3.6 27B0.041
  12. GLM-5.2 (max)0.0268
  13. Gemini 3.5 Flash0.0884
  14. GPT-5.5 (xhigh)0.107
  15. Claude Fable 5 (with fallback)0.367

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Benchmark interpretation
Higher values indicate stronger overall capability. Bars show how models compare on composite intelligence metrics.
Cost pressure
Ethen Gateway should route away from this model when the task is simple, repetitive, or likely to generate many output tokens.

Cost per Task

Weighted average cost (USD) per Intelligence Index task · Lower is better · Evaluation results measured independently by Artificial Analysis

10 models in the published order. Claude Opus 4.5 is not marked in this published set of 10 models.
  1. DeepSeek V4 Pro (max)$0.045
  2. gpt-oss-120b (high)$0.061
  3. MiniMax-M3$0.125
  4. Grok 4.3 (high)$0.139
  5. Nemotron 3 Ultra$0.244
  6. Kimi K2.6$0.346
  7. GLM-5.2 (max)$0.370
  8. Gemini 3.5 Flash$0.586
  9. GPT-5.5 (xhigh)$0.863
  10. Claude Fable 5 (with fallback)$2.75

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Cost pressure
Ethen Gateway should route away from this model when the task is simple, repetitive, or likely to generate many output tokens.

Cost to Run Artificial Analysis Intelligence Index

Cost (USD) to run all evaluations in the Artificial Analysis Intelligence Index · Evaluation results measured independently by Artificial Analysis

15 models in the published order. Claude Opus 4.5 is not marked in this published set of 15 models.
  1. gpt-oss-120b (high)3.5
  2. DeepSeek V3 03245.7
  3. DeepSeek V4 Pro (max)7.2
  4. GPT-5.5 (Non-reasoning)66.1
  5. Grok 4.3 (Non-reasoning)22.4
  6. MiniMax-M314.5
  7. Nova 2.0 Pro Preview53.3
  8. Grok 4.3 (high)12.5
  9. Qwen3.6 27B87.9
  10. Nemotron 3 Ultra33.5
  11. GLM-5.2 (max)32.9
  12. Kimi K2.625
  13. Gemini 3.5 Flash114
  14. GPT-5.5 (xhigh)131.8
  15. Claude Fable 5 (with fallback)507.6

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Benchmark interpretation
Higher values indicate stronger overall capability. Bars show how models compare on composite intelligence metrics.
Cost pressure
Ethen Gateway should route away from this model when the task is simple, repetitive, or likely to generate many output tokens.

Pricing: Cache Hit, Input, and Output

Price (USD per M Tokens) · Evaluation results measured independently by Artificial Analysis

18 models; segments: Cache Hit, Input, Output. Claude Opus 4.5: $30.50, row 15 of 18 in the published order.
  • Cache Hit
  • Input
  • Output
  1. gpt-oss-120b (high)$0.900
  2. DeepSeek V4 Pro (max)$1.31
  3. MiniMax-M3$1.56
  4. DeepSeek V3 0324$3.64
  5. Nemotron 3 Ultra$3.60
  6. Grok 4.3 (high)$3.95
  7. Grok 4.3 (Non-reasoning)$3.95
  8. Qwen3.6 27B$4.20
  9. Claude 3.5 Haiku$4.88
  10. Kimi K2.6$5.11
  11. GLM-5.2 (max)$6.06
  12. Gemini 3.5 Flash$10.65
  13. Nova 2.0 Pro Preview$11.25
  14. Claude Sonnet 5 (Non-reasoning)$12.20
  15. Claude Opus 4.5 (this model)$30.50
  16. GPT-5.5 (xhigh)$35.50
  17. GPT-5.5 (Non-reasoning)$35.50
  18. Claude Fable 5 (with fallback)$61.00

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Chart interpretation
Each bar is one model measured the same way, so this model can be read against the others. The chart lists its source and retrieval date.

Time per Intelligence Index Task

Weighted average decode time (minutes) per task; excludes TTFT and overhead time · Lower is better · Evaluation results measured independently by Artificial Analysis

14 models in the published order. Claude Opus 4.5 is not marked in this published set of 14 models.
  1. GPT-5.5 (Non-reasoning)0.351
  2. Nova 2.0 Pro Preview0.44
  3. Grok 4.3 (Non-reasoning)1.2
  4. Nemotron 3 Ultra1.4
  5. Grok 4.3 (high)1.6
  6. gpt-oss-120b (high)1.9
  7. Gemini 3.5 Flash2.2
  8. GLM-5.2 (max)2.8
  9. GPT-5.5 (xhigh)3
  10. Qwen3.6 27B3
  11. MiniMax-M34
  12. Claude Fable 5 (with fallback)4.8
  13. DeepSeek V4 Pro (max)6.7
  14. Kimi K2.68.2

Published benchmark dataset · Retrieved

Method and source
Methodology
Independent test run by Artificial Analysis on dedicated hardware.
Benchmark interpretation
Higher values indicate stronger overall capability. Bars show how models compare on composite intelligence metrics.

Specifications

Architecture and limits

Claude Opus 4.5 technical specifications
SpecificationValueSource
Model typeProprietaryfaq
ReasoningNofaq
Input modalitiesClaude Opus 4.5 (Non-reasoning) supports text and image input.faq
Output modalitiesClaude Opus 4.5 (Non-reasoning) supports text output.faq
Context window200k tokensfaq
Open weights / sourceNo, Claude Opus 4.5 (Non-reasoning) is proprietary. The model weights are not publicly available.faq
ParametersClaude Opus 4.5 (Non-reasoning) is a proprietary model and Anthropic has not disclosed the model size or parameter count.faq
Knowledge cutoffClaude Opus 4.5 (Non-reasoning) has a knowledge cutoff of August 2025. The model's training data includes information up to this date.faq
API availabilityYes, Claude Opus 4.5 (Non-reasoning) is available via API through 5 providers.faq

Method

Methodology & Provenance

This page is rendered from the normalized profile and page JSON for Claude Opus 4.5.

Benchmark values are preserved as normalized; only layout, disclosure ordering, and typography are adjusted for readability.

Questions

Model FAQ and technical disclosures

Claude Opus 4.5 (Non-reasoning) was released on November 24, 2025.

Claude Opus 4.5 (Non-reasoning) was created by Anthropic.

Claude Opus 4.5 (Non-reasoning) scores 35 (estimated) on the Artificial Analysis Intelligence Index, placing it well above average among other non-reasoning models in a similar price tier (median: 17).

Claude Opus 4.5 (Non-reasoning) generates output at 51.3 tokens per second (based on Anthropic's API), which is below average compared to other non-reasoning models in a similar price tier (median: 64.2 t/s).

Claude Opus 4.5 (Non-reasoning) has a time to first token (TTFT) of 1.51s (based on Anthropic's API), which is better than average compared to other non-reasoning models in a similar price tier (median: 1.56s).

Claude Opus 4.5 (Non-reasoning) costs $5.00 per 1M input tokens (at the higher end, median: $1.88) and $25.00 per 1M output tokens (at the higher end, median: $7.75), based on Anthropic's API.

Claude Opus 4.5 (Non-reasoning) costs $5.00 per 1M input tokens and $25.00 per 1M output tokens (based on Anthropic's API). For a blended rate (7:2:1 cache hit/input/output ratio), this is $3.85 per 1M tokens. Pricing may vary by provider.

No, Claude Opus 4.5 (Non-reasoning) is not a reasoning model. It provides direct responses without extended chain-of-thought reasoning.

Claude Opus 4.5 (Non-reasoning) supports text and image input.

Claude Opus 4.5 (Non-reasoning) supports text output.

Yes, Claude Opus 4.5 (Non-reasoning) supports image input and can analyze, describe, and answer questions about images.

Yes, Claude Opus 4.5 (Non-reasoning) is multimodal. It can process text and image input and generate text output.

Claude Opus 4.5 (Non-reasoning) has a context window of 200k tokens. This determines how much text and conversation history the model can process in a single request.

No, Claude Opus 4.5 (Non-reasoning) is proprietary. The model weights are not publicly available.

Claude Opus 4.5 (Non-reasoning) is a proprietary model and Anthropic has not disclosed the model size or parameter count.

Claude Opus 4.5 (Non-reasoning) achieves a score of 35 on the Artificial Analysis Intelligence Index. This composite benchmark evaluates models across reasoning, knowledge, mathematics, and coding.

Claude Opus 4.5 (Non-reasoning) has a knowledge cutoff of August 2025. The model's training data includes information up to this date.

Yes, Claude Opus 4.5 (Non-reasoning) is available via API through 5 providers.

Claude Opus 4.5 (Non-reasoning) is available through 5 API providers.

Models

  • 01AnthropicOther Anthropic models in the catalog.
  • 02LeaderboardsCertified metric ranks, when published.