Models / Claude Opus 5.5

Anthropic

Claude Opus 5.5

Claude Opus 5.5 is best for reasoning and knowledge, and agents and tool use; pick it when quality matters more than cost. The cheapest model of similar intelligence is Claude Sonnet 5.5, at $4.00 per million tokens against $8.00.

Results 16 Nov 2026 · 146 results · 11 sources

Test Claude Opus 5.5 on your prompt → It runs beside the models you could use instead, with every answer checked and costed.

Price per million tokens
$4.00 in · $20.00 out
Anthropic's list price, 10 Nov 2026 · compare prices$0.20 cached input · 50% off in batch
Speed
96 tokens a second; on its test prompt the answer starts after 10.3 min, thinking time included
Artificial Analysis, on its usual API, at max reasoning; 74–96 across 5 settings
Intelligence Index
57.6
Artificial Analysis, at max reasoning; 42.3–57.6 across 5 settings
Spring Prompt overall
2nd of 36 score 82 of 100 · how it works
Spring Prompt Score
77 of 100 5th of 32 · 4 of 5 benchmarks · our own tests only
Developer
Anthropic, based in the US
Weights not published
Released
22 Sep 2026
1,000,000 tokens of context

Where it stands among the models you could choose

Available models grouped by price, intelligence, speed and developer location. Retired models are excluded.

AmongIntelligence IndexSpring Prompt overallPrice, blendedSpeed
Models available today 1st of 185leads 2nd of 35Claude Fable 5.2 $8.00 blended, 92nd percentilecheapest: Mistral Nemo 37th of 65Nemotron 3.5 Super VL
Models priced $3–10 per million tokens 1st of 38leads 1st of 15leads sets this group 9th of 15Mistral Medium 3.5
Models writing 50–100 tokens a second 1st of 21leads 1st of 7leads $8.00 blended, 90th percentilecheapest: Granite 4.2 8B sets this group
Models from developers based in the US 1st of 92leads 2nd of 21Claude Fable 5.2 $8.00 blended, 83rd percentilecheapest: Gemma 3 4B 23rd of 33Nemotron 3.5 Super VL

The name under each rank is the leader of that group; green is the top quarter of the group and red the bottom quarter, by rank. Price is shown as a percentile: the share of the group that costs less, so lower is cheaper. "Sets this group" marks the measure the group is defined by. Spring Prompt overall is our score out of 100 across every source (how it works). Price is the developer's list price (or the typical OpenRouter host where we haven't read one) for three input tokens to one output; intelligence and speed are from Artificial Analysis (data sourced from Artificial Analysis); "similar intelligence" means within 4 points on its Intelligence Index. Developer locations are where each company is based, not where a model is served. Groups under 3 models are not ranked.

Measured cost per task

API cost recorded in our benchmark runs, excluding judging.

TaskBenchmarkCost per task
Product listingCatalogBench≈ $0.040
DeckDeckBench≈ $0.261
Survey plan and analysisSurveyBench≈ $0.457
Game of blitz chessBulletBench≈ $0.176

Cheapest model that passes

On our benchmarks that define a pass for each task: what one passing answer cost in our run, failed attempts included, and whether a cheaper model passed.

  • CatalogBench · cost per 1,000 accepted listings
    Claude Opus 5.5: $56.48 (119 of 168 attempts), 20 of 25. Cheapest: GPT-6 Luna at $0.70 (114 of 168 attempts). Cheaper and passed at least as often: GPT-6.1 Terra at $4.78 (125 of 168 attempts).
  • DeckBench · cost per presentable deck
    Claude Opus 5.5: $0.78 (2 of 6 tasks), 7 of 12. Cheapest: GPT-6.1 Terra at $0.13 (3 of 6 tasks). It passed at least as often.
  • SurveyBench · cost per sound survey task
    Claude Opus 5.5: $0.82 (5 of 9 tasks), 8 of 12. Cheapest: GPT-6 Luna at $0.03 (3 of 9 tasks). Cheaper and passed at least as often: GPT-6.1 Sol at $0.21 (6 of 9 tasks).

Cost per attempt divided by the share that passed, at the run date's prices, excluding judging. Each model at its cheapest setting per passing answer. With a handful of tasks, a pass or two either way is within chance. No estimate where a run had missing outputs, because their cost may not have been recorded.

How good is it, and for what?

Each use case ranks the available models its benchmarks measured, then those priced $3–10 per million tokens. The bar shows its percentile in that field, best to the right. Open a row for the results behind it.

Product listingsTurning a sparse product feed and photos into listings that can go live 13th of 32 available · 7th of 14 at its price
BenchmarkClaude Opus 5.5Rank among availableBest available
CatalogBenchCatalogBench: reliably publish-ready 55.4% 10th of 32 GPT-6 Astra 75.0%
CatalogBenchCatalogBench (sales brief): reliably publish-ready 21.4% 13th of 32 GPT-6 Astra 69.6%
Decks from an analysisTurning a finished analysis into a deck you could present as it is 7th of 32 available · 4th of 15 at its price
BenchmarkClaude Opus 5.5Rank among availableBest available
DeckBenchDeckBench: deck rating 1,323 7th of 32 Qwen4-Max 1,585
User surveysPlanning a user survey and reading its results without being misled 4th of 23 available · 3rd of 10 at its price
BenchmarkClaude Opus 5.5Rank among availableBest available
SurveyBenchSurveyBench: analysis rating 1,427 4th of 23 Claude Fable 5.2 1,641
Marketing planningPlanning a year of ad spend without overspending 9th of 32 available · 4th of 14 at its price
BenchmarkClaude Opus 5.5Rank among availableBest available
ROASBenchROASBench: overall score 51.7 9th of 32 Claude Fable 5.2 61.3
Agents and tool useMulti-step tasks with tools: support desks, coding agents, function calls 3rd of 159 available · 3rd of 35 at its price
BenchmarkClaude Opus 5.5Rank among availableBest available
Artificial AnalysisArtificial Analysis: Terminal-Bench 4.0 · reported 59.6% 3rd of 107 Kimi K3.1 Pro 67.2%
  • tau2-bench: pending. Its latest results are dated 26 May 2026, before this model was released.
  • Berkeley Function Calling Leaderboard (BFCL) V4: pending. Its latest results are dated 16 Dec 2025, before this model was released.
  • OpenHands Index: pending. Its latest results are dated 30 Jun 2026, before this model was released.
  • Microsoft STATE-Bench: pending. Its latest results are dated 29 May 2026, before this model was released.
  • Vending-Bench 2: has not measured this model.
Professional workReal tasks from banking, consulting and law, business documents and freelance projects 3rd of 51 available · 3rd of 23 at its price
BenchmarkClaude Opus 5.5Rank among availableBest available
APEX-AgentsAPEX-Agents: tasks passed · reported 73.5% 2nd of 37 Claude Sonnet 5.5 75.5%
GDP.pdfGDP.pdf: professional document tasks · reported 30.6% 2nd of 31 GPT-5.6 Sol 30.7%
  • Remote Labor Index: has not measured this model.
Reasoning and knowledgeHard questions across science, maths and general knowledge 1st of 187 available · 1st of 39 at its price
BenchmarkClaude Opus 5.5Rank among availableBest available
Artificial AnalysisArtificial Analysis: Artificial Analysis Intelligence Index · reported 57.6 1st of 185 This model
Artificial AnalysisArtificial Analysis: Humanity's Last Exam · reported 61.4% 1st of 186 This model
WritingWhat people prefer in blind comparisons, and judged writing quality 5th of 169 available · 3rd of 34 at its price
BenchmarkClaude Opus 5.5Rank among availableBest available
Arena (formerly LMArena)Arena Text: overall · reported 1,504 4th of 153 Gemini 4 Argon 1,525
UGI LeaderboardUGI: writing score · reported 70.0 8th of 124 Gemini 3.8 Flash 78.6
Sticking to the factsSummarising without inventing things, and factual answers 3rd of 139 available · 2nd of 31 at its price
BenchmarkClaude Opus 5.5Rank among availableBest available
Arena (formerly LMArena)Arena Text factuality: overall · reported 1,507 2nd of 109 Gemini 4 Argon 1,511
SimpleQA Verified (Epoch AI)SimpleQA Verified: correct answers · reported 72.2% 4th of 68 GPT-6 Astra 75.6%
  • Vectara Hallucination Leaderboard: has not measured this model.
Speed under pressureGood decisions against a real clock (fast chess) Too slow for the clock
  • BulletBench: too slow for Bullet 60s: lost its first 4 games on time, at a median 3.1 s a move; too slow for Lightning 10+1: lost its first 4 games on time, at a median 3.0 s a move.

Nearest alternatives

Against Claude Opus 5.5 at $8.00 blended, Intelligence Index 57.6. "Similar intelligence" means within 4 points or better.

Where to run it

10 hosts on OpenRouter at the standard service tier, with a median of $8.40 blended per million tokens. Cheapest first.

HostQuantisationInputOutputBlended
Amazon Bedrocknot stated$4.00$20.00$8.00
Anthropicnot stated$4.00$20.00$8.00
Azurenot stated$4.00$20.00$8.00
Claude Platform on AWSnot stated$4.00$20.00$8.00
Googlenot stated$4.00$20.00$8.00
Amazon Bedrocknot stated$4.40$22.00$8.80
Amazon Bedrocknot stated$4.40$22.00$8.80
Azurenot stated$4.40$22.00$8.80

Other service tiers, not counted above: Anthropic fast $8.00 in, $40.00 out. Flex and batch are cheaper but slower or queued; priority and fast cost more.

Prices per million tokens from OpenRouter's public endpoint list, checked 10 Nov 2026; the 2 dearer hosts are left out.

Against the alternatives

The models you would most likely weigh it against: the leaders of the groups above.

Scroll sideways to see every alternative.

Claude Opus 5.5Claude Fable 5.2
highest-scoring overall available today
GPT-6.1 Sol
just below overall
Price per million tokens $8.00 $16.00$4.00
Tokens a second 96 –53
Intelligence Index 57.6 –51.8
Spring Prompt overall 82 9080
Rank among available models, by use case
Product listings 13th of 32 5th2nd
Decks from an analysis 7th of 32 4th5th
User surveys 4th of 23 1st3rd
Marketing planning 9th of 32 1st–
Agents and tool use 3rd of 159 –6th
Professional work 3rd of 51 –9th
Reasoning and knowledge 1st of 187 –6th
Writing 5th of 169 –16th
Sticking to the facts 3rd of 139 –18th
Speed under pressure 36th of 36 8th3rd

Shaded figures are better than Claude Opus 5.5's. A dash means no result.

Research and examples

In the playground: our pick in 14 of 29 tests, 93% of checks passed →

Where it has been measured

Most intelligent of the 92 models from developers based in the US, and highest-scoring overall of the 15 models priced $3–10 per million tokens. Most intelligent of the 38 models priced $3–10 per million tokens. Among models available today, in the top quarter (by percentile) for 7 of the 10 use cases measured, led by reasoning and knowledge, and agents and tool use; in the bottom quarter for speed under pressure.

Every result

146 published results from 11 sources, each in the source's own units, with the configuration that produced it.

BulletBench · measured by Spring Prompt · 5 results
MeasureValueRankConfigurationDated
BulletBench Blitz 3+2: cost per gameUS dollars, lower is better $0.18 16 of 22 Claude Opus 5.5
provider default reasoning
29 Sep 2026
BulletBench Blitz 3+2: games lost on time% of games, lower is better 62.5% 18 of 22 Claude Opus 5.5
provider default reasoning
29 Sep 2026
BulletBench Blitz 3+2: invalid moves% of moves, lower is better 0.0% 1 of 22 Claude Opus 5.5
provider default reasoning
29 Sep 2026
BulletBench Blitz 3+2: ladder Eloladder Elo, higher is better 557
370–727
12 of 22 Claude Opus 5.5
provider default reasoning
29 Sep 2026
BulletBench Blitz 3+2: median move timemilliseconds, lower is better 7.5 s 20 of 22 Claude Opus 5.5
provider default reasoning
29 Sep 2026

Not shown: Not your cost: prices are those charged on the run date.

CatalogBench · measured by Spring Prompt · 31 results
MeasureValueRankConfigurationDated
CatalogBench (sales brief): channel rules broken% of products, lower is better 0.6% 17 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench (sales brief): claims to check% of products, lower is better 0.0% 1 of 32 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench (sales brief): content quality% of checks, higher is better 97.8% 13 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench (sales brief): failed outputs% of products, lower is better 0.0% 1 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench (sales brief): missing UK information% of products, lower is better 0.0% 1 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench (sales brief): not findable% of products, lower is better 1.8% 5 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench (sales brief): publish-ready listings% of products, higher is better 34.5% 15 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench (sales brief): reliably publish-ready% of products, higher is better 21.4% 13 of 34 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench (sales brief): unsupported claims% of products, lower is better 60.1% 19 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench (sales brief): unsupported claimsclaims per product, lower is better 1.66 17 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench (sales brief): wrong attributes% of products, lower is better 15.5% 13 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench (sales brief): wrong category or variant% of products, lower is better 1.8% 6 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench: channel compliance% of products, higher is better 100.0% 1 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench: channel rules broken% of products, lower is better 0.0% 1 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench: claims to check% of products, lower is better 0.0% 1 of 32 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench: conflicts caught% of conflicts, higher is better 100.0% 1 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench: content quality% of checks, higher is better 97.9% 7 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench: cost per productUS dollars, lower is better $0.0400 31 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench: decision accuracy% of decisions, higher is better 97.7% 9 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench: failed outputs% of products, lower is better 0.0% 1 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench: field accuracy% of missing fields, higher is better 92.2% 28 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench: invented values% of filled values, lower is better 0.5% 6 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench: missing UK information% of products, lower is better 0.0% 1 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench: not findable% of products, lower is better 3.0% 9 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench: publish-ready listings% of products, higher is better 70.8% 10 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench: reliably publish-ready% of products, higher is better 55.4% 10 of 34 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench: time per productseconds, lower is better 16.9 16 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench: unsupported claims% of products, lower is better 17.3% 18 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench: unsupported claimsclaims per product, lower is better 0.45 16 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench: wrong attributes% of products, lower is better 15.5% 11 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026
CatalogBench: wrong category or variant% of products, lower is better 1.8% 6 of 35 Claude Opus 5.5
provider default reasoning
30 Sep 2026

Not shown: Not your catalogue: invented products with generated images, so a model's result on your own feed can differ.

DeckBench · measured by Spring Prompt · 17 results
MeasureValueRankConfigurationDated
DeckBench: accurate decks% of tasks, higher is better 83.3% 5 of 34 Claude Opus 5.5
provider default reasoning
6 Oct 2026
DeckBench: caveat dropped% of tasks, lower is better 16.7% 15 of 34 Claude Opus 5.5
provider default reasoning
6 Oct 2026
DeckBench: clean layout% of tasks, higher is better 83.3% 3 of 34 Claude Opus 5.5
provider default reasoning
6 Oct 2026
DeckBench: cost per deckUS dollars, lower is better $0.26 29 of 34 Claude Opus 5.5
provider default reasoning
6 Oct 2026
DeckBench: deck ratingrating, higher is better 1,323
1,247–1,403
7 of 34 Claude Opus 5.5
provider default reasoning
6 Oct 2026
DeckBench: design quality% of the maximum, higher is better 72.0% 11 of 34 Claude Opus 5.5
provider default reasoning
6 Oct 2026
DeckBench: draft figure quoted% of tasks, lower is better 0.0% 1 of 34 Claude Opus 5.5
provider default reasoning
6 Oct 2026
DeckBench: findings missing% of tasks, lower is better 0.0% 1 of 34 Claude Opus 5.5
provider default reasoning
6 Oct 2026
DeckBench: head-to-head win rate% of comparisons, higher is better 79.5% 6 of 34 Claude Opus 5.5
provider default reasoning
6 Oct 2026
DeckBench: layout defects% of tasks, lower is better 16.7% 3 of 34 Claude Opus 5.5
provider default reasoning
6 Oct 2026
DeckBench: misleading metric used% of tasks, lower is better 0.0% 1 of 34 Claude Opus 5.5
provider default reasoning
6 Oct 2026
DeckBench: presentable decks% of tasks, higher is better 33.3% 7 of 34 Claude Opus 5.5
provider default reasoning
6 Oct 2026
DeckBench: recommendation late or wrong% of tasks, lower is better 0.0% 1 of 34 Claude Opus 5.5
provider default reasoning
6 Oct 2026
DeckBench: slides needing work% of tasks, lower is better 66.7% 24 of 34 Claude Opus 5.5
provider default reasoning
6 Oct 2026
DeckBench: time per deckseconds, lower is better 98.5 14 of 34 Claude Opus 5.5
provider default reasoning
6 Oct 2026
DeckBench: unsupported claims% of tasks, lower is better 0.0% 1 of 34 Claude Opus 5.5
provider default reasoning
6 Oct 2026
DeckBench: unsupported numbers% of tasks, lower is better 0.0% 1 of 34 Claude Opus 5.5
provider default reasoning
6 Oct 2026

Not shown: Not your analysis or your brand: invented companies, six public tasks, one deck per model per task at the provider's default reasoning setting.

ROASBench · measured by Spring Prompt · 9 results
MeasureValueRankConfigurationDated
ROASBench: audience scorescore out of 100, higher is better 67.7 6 of 35 Claude Opus 5.5
provider default reasoning
29 Sep 2026
ROASBench: business scorescore out of 100, higher is better 52.3 9 of 35 Claude Opus 5.5
provider default reasoning
29 Sep 2026
ROASBench: consistency scorescore out of 100, higher is better 33.8 11 of 35 Claude Opus 5.5
provider default reasoning
29 Sep 2026
ROASBench: contribution profitsimulated US dollars, higher is better $911,566 3 of 35 Claude Opus 5.5
provider default reasoning
29 Sep 2026
ROASBench: cost of a runUS dollars, lower is better $1.06 31 of 35 Claude Opus 5.5
provider default reasoning
29 Sep 2026
ROASBench: months over budgetmonths of 12, lower is better 0 1 of 35 Claude Opus 5.5
provider default reasoning
29 Sep 2026
ROASBench: overall scorescore out of 100, higher is better 51.7 9 of 35 Claude Opus 5.5
provider default reasoning
29 Sep 2026
ROASBench: planning scorescore out of 100, higher is better 54.7 22 of 35 Claude Opus 5.5
provider default reasoning
29 Sep 2026
ROASBench: return on ad spendprofit per $1 spent, higher is better $0.69 6 of 35 Claude Opus 5.5
provider default reasoning
29 Sep 2026

Not shown: Not real ad performance: the market is a deterministic simulation of one invented skincare brand.

SurveyBench · measured by Spring Prompt · 20 results
MeasureValueRankConfigurationDated
SurveyBench: analysis ratingrating, higher is better 1,427
1,376–1,486
4 of 25 Claude Opus 5.5
provider default reasoning
8 Oct 2026
SurveyBench: analysis scorepoints of 100, higher is better 95.3 4 of 25 Claude Opus 5.5
provider default reasoning
8 Oct 2026
SurveyBench: cost per taskUS dollars, lower is better $0.46 22 of 24 Claude Opus 5.5
provider default reasoning
8 Oct 2026
SurveyBench: head-to-head win rate% of comparisons, higher is better 76.2% 3 of 25 Claude Opus 5.5
provider default reasoning
8 Oct 2026
SurveyBench: leading or double-barrelled questions% of tasks, lower is better 0.0% 1 of 25 Claude Opus 5.5
provider default reasoning
8 Oct 2026
SurveyBench: mistaken next steps% of tasks, lower is better 0.0% 1 of 25 Claude Opus 5.5
provider default reasoning
8 Oct 2026
SurveyBench: numbers right% of answers, higher is better 100.0% 1 of 25 Claude Opus 5.5
provider default reasoning
8 Oct 2026
SurveyBench: objectives unanswerable% of tasks, lower is better 0.0% 1 of 25 Claude Opus 5.5
provider default reasoning
8 Oct 2026
SurveyBench: precision ignored% of tasks, lower is better 0.0% 1 of 25 Claude Opus 5.5
provider default reasoning
8 Oct 2026
SurveyBench: research scorepoints of 100, higher is better 96.1 3 of 25 Claude Opus 5.5
provider default reasoning
8 Oct 2026
SurveyBench: sound research% of tasks, higher is better 55.6% 2 of 25 Claude Opus 5.5
provider default reasoning
8 Oct 2026
SurveyBench: stakeholder's view presumed% of tasks, lower is better 0.0% 1 of 25 Claude Opus 5.5
provider default reasoning
8 Oct 2026
SurveyBench: survey plan scorepoints of 100, higher is better 97.8 1 of 25 Claude Opus 5.5
provider default reasoning
8 Oct 2026
SurveyBench: survey structure% of tasks, lower is better 11.1% 3 of 25 Claude Opus 5.5
provider default reasoning
8 Oct 2026
SurveyBench: time per taskseconds, lower is better 180 13 of 24 Claude Opus 5.5
provider default reasoning
8 Oct 2026
SurveyBench: traps handled% of traps, higher is better 95.7% 5 of 25 Claude Opus 5.5
provider default reasoning
8 Oct 2026
SurveyBench: traps missed% of tasks, lower is better 33.3% 6 of 25 Claude Opus 5.5
provider default reasoning
8 Oct 2026
SurveyBench: unsupported findings% of tasks, lower is better 22.2% 5 of 25 Claude Opus 5.5
provider default reasoning
8 Oct 2026
SurveyBench: wrong numbers% of tasks, lower is better 0.0% 1 of 25 Claude Opus 5.5
provider default reasoning
8 Oct 2026
SurveyBench: wrong recommendation% of tasks, lower is better 0.0% 1 of 25 Claude Opus 5.5
provider default reasoning
8 Oct 2026

Not shown: Not your customers or your data: invented organisations with simulated respondents, nine public tasks, one run per model per task at the provider's default reasoning setting.

APEX-Agents · reported by APEX-Agents · 1 result
MeasureValueRankConfigurationDated
APEX-Agents: tasks passed% of tasks, higher is better 73.5% 2 of 39 Claude Opus 5.5 (max reasoning) 1 Oct 2026

Not shown: Not your firm's documents or tools; graded by rubric, not by a client.

Arena (formerly LMArena) · reported by Arena (formerly LMArena) · 11 results
MeasureValueRankConfigurationDated
Arena Agent: confirmed task successIPS effect estimate, higher is better 0.14
0.10–0.18
4 of 50 Claude Opus 5.5 (high reasoning, arena agent) 2 Oct 2026
Arena Agent: praise over complaintIPS effect estimate, higher is better 0.31
0.23–0.40
3 of 50 Claude Opus 5.5 (high reasoning, arena agent) 2 Oct 2026
Arena Agent: steerabilityIPS effect estimate, higher is better 0.10
0.08–0.13
4 of 50 Claude Opus 5.5 (high reasoning, arena agent) 2 Oct 2026
Arena Agent: tool groundingIPS effect estimate, higher is better 0.00
0.00–0.01
6 of 50 Claude Opus 5.5 (high reasoning, arena agent) 2 Oct 2026
Arena Text factuality: overallArena rating, higher is better 1,507
1,499–1,515
2 of 182 Claude Opus 5.5 (high reasoning) 2 Oct 2026
Arena Text: business, management and financeArena rating, higher is better 1,489
1,468–1,510
12 of 406 Claude Opus 5.5 (high reasoning) 2 Oct 2026
Arena Text: creative writingArena rating, higher is better 1,516
1,497–1,536
2 of 411 Claude Opus 5.5 (high reasoning) 2 Oct 2026
Arena Text: expert promptsArena rating, higher is better 1,541
1,514–1,567
6 of 363 Claude Opus 5.5 (high reasoning) 2 Oct 2026
Arena Text: instruction followingArena rating, higher is better 1,517
1,502–1,533
2 of 413 Claude Opus 5.5 (high reasoning) 2 Oct 2026
Arena Text: overallArena rating, higher is better 1,504
1,495–1,513
4 of 413 Claude Opus 5.5 (high reasoning) 2 Oct 2026
Arena Text: writing, literature and languageArena rating, higher is better 1,519
1,502–1,536
2 of 412 Claude Opus 5.5 (high reasoning) 2 Oct 2026

Not shown: Not a success rate: the estimated effect of choosing this model as the agent's orchestrator, against an even mix of the models tested, inside Arena's own harness.

Artificial Analysis · reported by Artificial Analysis · 35 results
MeasureValueRankConfigurationDated
Artificial Analysis: Artificial Analysis Intelligence Indexindex score, higher is better 53.6 4 of 413 Claude Opus 5.5 (high reasoning) 20 Oct 2026
Artificial Analysis: Artificial Analysis Intelligence Indexindex score, higher is better 42.3 48 of 413 Claude Opus 5.5 (low reasoning) 20 Oct 2026
Artificial Analysis: Artificial Analysis Intelligence Indexindex score, higher is better 57.6 1 of 413 Claude Opus 5.5 (max reasoning) 20 Oct 2026
Artificial Analysis: Artificial Analysis Intelligence Indexindex score, higher is better 51.2 13 of 413 Claude Opus 5.5 (medium reasoning) 20 Oct 2026
Artificial Analysis: Artificial Analysis Intelligence Indexindex score, higher is better 56.0 2 of 413 Claude Opus 5.5 (xhigh reasoning) 20 Oct 2026
Artificial Analysis: Humanity's Last Exam% of questions, higher is better 55.6% 7 of 419 Claude Opus 5.5 (high reasoning) 20 Oct 2026
Artificial Analysis: Humanity's Last Exam% of questions, higher is better 48.3% 33 of 419 Claude Opus 5.5 (low reasoning) 20 Oct 2026
Artificial Analysis: Humanity's Last Exam% of questions, higher is better 61.4% 1 of 419 Claude Opus 5.5 (max reasoning) 20 Oct 2026
Artificial Analysis: Humanity's Last Exam% of questions, higher is better 54.7% 11 of 419 Claude Opus 5.5 (medium reasoning) 20 Oct 2026
Artificial Analysis: Humanity's Last Exam% of questions, higher is better 57.5% 4 of 419 Claude Opus 5.5 (xhigh reasoning) 20 Oct 2026
Artificial Analysis: Long-context reasoning (AA-LCR)% of questions, higher is better 82.7% 32 of 407 Claude Opus 5.5 (high reasoning) 20 Oct 2026
Artificial Analysis: Long-context reasoning (AA-LCR)% of questions, higher is better 80.7% 64 of 407 Claude Opus 5.5 (low reasoning) 20 Oct 2026
Artificial Analysis: Long-context reasoning (AA-LCR)% of questions, higher is better 84.7% 5 of 407 Claude Opus 5.5 (max reasoning) 20 Oct 2026
Artificial Analysis: Long-context reasoning (AA-LCR)% of questions, higher is better 84.3% 8 of 407 Claude Opus 5.5 (medium reasoning) 20 Oct 2026
Artificial Analysis: Long-context reasoning (AA-LCR)% of questions, higher is better 84.7% 5 of 407 Claude Opus 5.5 (xhigh reasoning) 20 Oct 2026
Artificial Analysis: Output speedtokens per second, higher is better 75.2 90 of 137 Claude Opus 5.5 (high reasoning) 20 Oct 2026
Artificial Analysis: Output speedtokens per second, higher is better 75.6 89 of 137 Claude Opus 5.5 (low reasoning) 20 Oct 2026
Artificial Analysis: Output speedtokens per second, higher is better 96.0 72 of 137 Claude Opus 5.5 (max reasoning) 20 Oct 2026
Artificial Analysis: Output speedtokens per second, higher is better 74.0 92 of 137 Claude Opus 5.5 (medium reasoning) 20 Oct 2026
Artificial Analysis: Output speedtokens per second, higher is better 80.3 83 of 137 Claude Opus 5.5 (xhigh reasoning) 20 Oct 2026
Artificial Analysis: SciCode% of problems, higher is better 60.4% 10 of 205 Claude Opus 5.5 (high reasoning) 20 Oct 2026
Artificial Analysis: SciCode% of problems, higher is better 58.6% 21 of 205 Claude Opus 5.5 (low reasoning) 20 Oct 2026
Artificial Analysis: SciCode% of problems, higher is better 66.9% 1 of 205 Claude Opus 5.5 (max reasoning) 20 Oct 2026
Artificial Analysis: SciCode% of problems, higher is better 59.3% 15 of 205 Claude Opus 5.5 (medium reasoning) 20 Oct 2026
Artificial Analysis: SciCode% of problems, higher is better 65.0% 2 of 205 Claude Opus 5.5 (xhigh reasoning) 20 Oct 2026
Artificial Analysis: Terminal-Bench 4.0% of tasks, higher is better 56.6% 9 of 204 Claude Opus 5.5 (high reasoning) 20 Oct 2026
Artificial Analysis: Terminal-Bench 4.0% of tasks, higher is better 31.3% 46 of 204 Claude Opus 5.5 (low reasoning) 20 Oct 2026
Artificial Analysis: Terminal-Bench 4.0% of tasks, higher is better 59.6% 3 of 204 Claude Opus 5.5 (max reasoning) 20 Oct 2026
Artificial Analysis: Terminal-Bench 4.0% of tasks, higher is better 52.5% 15 of 204 Claude Opus 5.5 (medium reasoning) 20 Oct 2026
Artificial Analysis: Terminal-Bench 4.0% of tasks, higher is better 59.6% 3 of 204 Claude Opus 5.5 (xhigh reasoning) 20 Oct 2026
Artificial Analysis: Time to first answer tokenseconds, lower is better 26.6 92 of 137 Claude Opus 5.5 (high reasoning) 20 Oct 2026
Artificial Analysis: Time to first answer tokenseconds, lower is better 6.89 50 of 137 Claude Opus 5.5 (low reasoning) 20 Oct 2026
Artificial Analysis: Time to first answer tokenseconds, lower is better 621 137 of 137 Claude Opus 5.5 (max reasoning) 20 Oct 2026
Artificial Analysis: Time to first answer tokenseconds, lower is better 13.8 69 of 137 Claude Opus 5.5 (medium reasoning) 20 Oct 2026
Artificial Analysis: Time to first answer tokenseconds, lower is better 80.9 126 of 137 Claude Opus 5.5 (xhigh reasoning) 20 Oct 2026

Not shown: Not business work, and a blend: read the parts for any one task.

GDP.pdf · reported by GDP.pdf · 1 result
MeasureValueRankConfigurationDated
GDP.pdf: professional document tasks% of rubric, higher is better 30.6% 3 of 49 Claude Opus 5.5 (max reasoning) 1 Oct 2026

Not shown: Not your documents; 100 tasks, so small differences are noise.

SimpleQA Verified (Epoch AI) · reported by SimpleQA Verified (Epoch AI) · 1 result
MeasureValueRankConfigurationDated
SimpleQA Verified: correct answers% of questions, higher is better 72.2% 4 of 85 Claude Opus 5.5 (max reasoning) 22 Sep 2026

Not shown: Not answers grounded in your documents; tests what the model remembers.

UGI Leaderboard · reported by UGI Leaderboard · 15 results
MeasureValueRankConfigurationDated
UGI: requested-length error% off the requested word count, lower is better 5.0% 41 of 379 Claude Opus 5.5 (high reasoning) 23 Sep 2026
UGI: requested-length error% off the requested word count, lower is better 5.0% 41 of 379 Claude Opus 5.5 (low reasoning) 23 Sep 2026
UGI: requested-length error% off the requested word count, lower is better 2.0% 10 of 379 Claude Opus 5.5 (max reasoning) 23 Sep 2026
UGI: requested-length error% off the requested word count, lower is better 5.0% 41 of 379 Claude Opus 5.5 (medium reasoning) 23 Sep 2026
UGI: requested-length error% off the requested word count, lower is better 5.0% 41 of 379 Claude Opus 5.5 (xhigh reasoning) 23 Sep 2026
UGI: style adherencescore from 0 to 1, higher is better 0.43 4 of 379 Claude Opus 5.5 (high reasoning) 23 Sep 2026
UGI: style adherencescore from 0 to 1, higher is better 0.40 22 of 379 Claude Opus 5.5 (low reasoning) 23 Sep 2026
UGI: style adherencescore from 0 to 1, higher is better 0.43 1 of 379 Claude Opus 5.5 (max reasoning) 23 Sep 2026
UGI: style adherencescore from 0 to 1, higher is better 0.43 2 of 379 Claude Opus 5.5 (medium reasoning) 23 Sep 2026
UGI: style adherencescore from 0 to 1, higher is better 0.42 5 of 379 Claude Opus 5.5 (xhigh reasoning) 23 Sep 2026
UGI: writing scorescore out of 100, higher is better 68.4 35 of 379 Claude Opus 5.5 (high reasoning) 23 Sep 2026
UGI: writing scorescore out of 100, higher is better 70.0 22 of 379 Claude Opus 5.5 (low reasoning) 23 Sep 2026
UGI: writing scorescore out of 100, higher is better 68.9 31 of 379 Claude Opus 5.5 (max reasoning) 23 Sep 2026
UGI: writing scorescore out of 100, higher is better 69.8 24 of 379 Claude Opus 5.5 (medium reasoning) 23 Sep 2026
UGI: writing scorescore out of 100, higher is better 69.4 28 of 379 Claude Opus 5.5 (xhigh reasoning) 23 Sep 2026

Not shown: Not other format limits such as character counts or bullet counts.

Embed a score badge

A badge for its Spring Prompt Score and each of our benchmarks that ranked Claude Opus 5.5, from the current results. Paste the code into a page or README; it links back to where the result comes from.

Claude Opus 5.5 is #5 of 32 on the Spring Prompt Score (77 of 100, from 4 of 5 of our benchmarks), measured by Spring Prompt Spring Prompt Score · 4 of 5 benchmarks
<a href="https://springprompt.com/score#anthropic-claude-opus-5-5"><img src="https://springprompt.com/badge/score/anthropic-claude-opus-5-5.svg" alt="Claude Opus 5.5 is #5 of 32 on the Spring Prompt Score (77 of 100, from 4 of 5 of our benchmarks), measured by Spring Prompt" height="20"></a>
Claude Opus 5.5 is #10 of 32 on CatalogBench (Reliably publish-ready, factual brief: 55.4%), measured by Spring Prompt CatalogBench · Reliably publish-ready, factual brief
<a href="https://springprompt.com/benchmarks/catalogbench?focus=anthropic-claude-opus-5-5"><img src="https://springprompt.com/badge/catalogbench/anthropic-claude-opus-5-5.svg" alt="Claude Opus 5.5 is #10 of 32 on CatalogBench (Reliably publish-ready, factual brief: 55.4%), measured by Spring Prompt" height="20"></a>
Claude Opus 5.5 is #7 of 33 on DeckBench (Deck rating: 1,323), measured by Spring Prompt DeckBench · Deck rating
<a href="https://springprompt.com/benchmarks/deckbench?focus=anthropic-claude-opus-5-5"><img src="https://springprompt.com/badge/deckbench/anthropic-claude-opus-5-5.svg" alt="Claude Opus 5.5 is #7 of 33 on DeckBench (Deck rating: 1,323), measured by Spring Prompt" height="20"></a>
Claude Opus 5.5 is #4 of 24 on SurveyBench (Analysis rating: 1,427), measured by Spring Prompt SurveyBench · Analysis rating
<a href="https://springprompt.com/benchmarks/surveybench?focus=anthropic-claude-opus-5-5"><img src="https://springprompt.com/badge/surveybench/anthropic-claude-opus-5-5.svg" alt="Claude Opus 5.5 is #4 of 24 on SurveyBench (Analysis rating: 1,427), measured by Spring Prompt" height="20"></a>
Claude Opus 5.5 is #9 of 33 on ROASBench (Score: 51.7), measured by Spring Prompt ROASBench · Score
<a href="https://springprompt.com/benchmarks/roasbench?focus=anthropic-claude-opus-5-5"><img src="https://springprompt.com/badge/roasbench/anthropic-claude-opus-5-5.svg" alt="Claude Opus 5.5 is #9 of 33 on ROASBench (Score: 51.7), measured by Spring Prompt" height="20"></a>
Claude Opus 5.5 is #11 of 20 on BulletBench (Ladder Elo: 557), measured by Spring Prompt BulletBench · Ladder Elo
<a href="https://springprompt.com/benchmarks/bulletbench?focus=anthropic-claude-opus-5-5"><img src="https://springprompt.com/badge/bulletbench/anthropic-claude-opus-5-5.svg" alt="Claude Opus 5.5 is #11 of 20 on BulletBench (Ladder Elo: 557), measured by Spring Prompt" height="20"></a>

Will Claude Opus 5.5 work for your product?

Benchmarks aren't your task. Paste the prompt your product uses and run it beside the models you could use instead, with every answer checked and costed.

New here? Your email starts a free trial. No credit card needed.