Claude Opus 4.6 (max): benchmark scores, price and ranking

Claude Opus 4.6 (max) from Anthropic ranks #63 of 743 LLMs on the LLMs Tiger aggregate benchmark score (52.5), based on 16 benchmark results captured up to 2026-10-13.

Organisation

Anthropic

Type

Frontier model

Open weights

No

Released

2026-02-05 (source: curated)

Cheapest API price

$5.00 in / $25.00 out per million tokens (via anthropic, observed 2026-10-13T19:47Z)

Aggregate rank

#63 of 743 (score 52.5)

Relative to other models, Claude Opus 4.6 (max) is strongest on LMCA (#6 of 108) and weakest on Furniture Assembly (#23 of 28).

Across the 16 benchmarks where at least ten models are scored, Claude Opus 4.6 (max) is in the top 10% on 1 and in the top half on 12.

Its cheapest listed input price, $5.00 per million tokens, is lower than that of 7% of the 427 models with a listed price.

Claude Opus 4.6 (max) benchmark results

About Claude Opus 4.6 (max)

Claude Opus 4.6 (max) is a frontier model from Anthropic. It was released on 2026-02-05, according to the developer's announcement. The cheapest API listing we track charges $5.00 per million input tokens and $25.00 per million output tokens, through anthropic. We hold 16 benchmark results for it across 3 areas, and 1 is in the top tenth of its board.

Where Claude Opus 4.6 (max) stands

On our aggregate score, Claude Opus 4.6 (max) is #63 of 743 ranked models, in the top quarter. Its score is 52.5; the scale is centred on 50, so it sits 2.5 points above the typical model. That beats 92% of the ranked field. Among the 70 ranked Anthropic models it is #17. The next one up is claude-opus-4-20250514 at 52.5. The next one down is Claude Haiku 5.5 (max) at 52.1.

Claude Opus 4.6 (max), benchmark by benchmark

On LMCA, Claude Opus 4.6 (max) scores 55.8% and ranks #6 of 108. Only 5 models score higher, led by Claude Opus 5.5 (max) at 68.2%, which is 12.5 points ahead. It covers language-model capability assessment. The board median is 29.2%, so Claude Opus 4.6 (max) is 26.5 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

On DTBench, Claude Opus 4.6 (max) reaches 91.2%. It ranks #18 of 146, ahead of 88% of the field, and 7.7 points behind Claude Opus 5.5 (max) at 98.9%. The results either side are Qwen3.8 Max (xhigh) (92.0%) and Kimi K3 (max) (91.2%). It covers dynamic tool-use benchmark. The board median is 68.8%, so Claude Opus 4.6 (max) is 22.4 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

Claude Opus 4.6 (max) manages 34.4% on Humanity's Last Exam, good for #4 of 27. Muse Spark leads at 40.6%, 6.1 points further up, and 85% of models sit below. The results either side are GPT-5.4 (xhigh) (36.2%) and GPT-5 (high) (25.3%). It covers expert-level questions across domains. The board median is 10.7%, so Claude Opus 4.6 (max) is 23.7 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

GPQA Diamond puts Claude Opus 4.6 (max) at 88.4%, #46 of 275. That beats 83% of the models tested, but 7.4 points short of GPT-6 Astra (max) (95.8%). The results either side are Qwen3.5 397B-A17B (88.4%) and Qwen 3.6 Plus (2026-04-02) (88.4%). It covers expert-level science questions. The board median is 71.7%, so Claude Opus 4.6 (max) is 16.7 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

On OTIS Mock AIME, Claude Opus 4.6 (max) reaches 91.1%. It ranks #39 of 189, ahead of 79% of the field, and 8.9 points behind Claude Fable 5.1 (max) at 100.0%. The results either side are GPT-5 (high) (91.4%) and GLM-5.3 (max) (91.1%). It covers competition maths (AIME 2024-25 style). The board median is 57.8%, so Claude Opus 4.6 (max) is 33.3 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

Claude Opus 4.6 (max) manages 7.6% on EnigmaEval, good for #5 of 24. Gemini 3 Pro Preview leads at 18.2%, 10.6 points further up, and 79% of models sit below. The results either side are o4-mini (high) (9.2%) and Claude Opus 4.1 (unknown thinking) (7.2%). It covers multi-step puzzle hunts. The board median is 2.9%, so Claude Opus 4.6 (max) is 4.7 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

Claude Opus 4.6 (max) lands below the middle on FrontierMath T4: 26.8%, #28 of 50, with 54% of the field scoring higher. GPT-6.1 Sol (max) leads at 100.0%; the closest result above is GLM-5.3 (max) at 29.3%. It covers research-level maths, tier 4. The board median is 30.5%, so Claude Opus 4.6 (max) is 3.7 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

Claude Opus 4.6 (max) lands below the middle on APEX-Agents: 46.3%, #9 of 16, with 50% of the field scoring higher. Claude Sonnet 5.5 (max) leads at 75.5%; the closest result above is Claude Opus 4.7 (max) at 49.2%. It covers long professional agent tasks. The board median is 47.8%, so Claude Opus 4.6 (max) is 1.5 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

EBR-Bench is a weak spot. Claude Opus 4.6 (max) scores 12.7%, #17 of 21, behind 76% of the models measured. The leader, GPT-6 Astra (max), reaches 76.2%. It covers evidence-based reasoning. The board median is 31.4%, so Claude Opus 4.6 (max) is 18.7 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

Furniture Assembly is a weak spot. Claude Opus 4.6 (max) scores 28.3%, #23 of 28, behind 79% of the models measured. The leader, Claude Opus 5.5 (max), reaches 83.3%. It covers planning a physical assembly. The board median is 41.2%, so Claude Opus 4.6 (max) is 12.9 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.

The full table above also lists the other 6 results, including boards with fewer than ten models, where a rank says little.

Strengths and weaknesses

Reasoning, maths and science: solid, ahead of 72% of the field on average across 7 benchmarks. Its best board here is Humanity's Last Exam (#4 of 27) and its weakest is FrontierMath T4 (#28 of 50). Agents and tool use: solid, ahead of 68% of the field on average across 3 benchmarks. Its best board here is DTBench (#18 of 146) and its weakest is APEX-Agents (#9 of 16). General knowledge and chat: solid, ahead of 50% of the field on average across 6 benchmarks. Its best board here is LMCA (#6 of 108) and its weakest is Furniture Assembly (#23 of 28).

Price and value

The median input price across the 427 priced models is $0.500 per million tokens, so Claude Opus 4.6 (max) costs more than the typical listing. Models that score within 1.5 points or better and cost less: GLM-5.2 (high) ($0.060, 51.6), DeepSeek V4 Flash 0731 (max) ($0.070, 53.1), Qwen3-Coder-30B-A3B-Instruct ($0.070, 52.7). Similar scores cost more on Claude Mythos Preview (Early) ($10.00) and o1-2024-12-17 ($15.00).

Release history

The Anthropic model released before it was Claude Opus 4.6 (120K, Medium) on 2026-02-05, which scores 51.3. The next one was Claude Sonnet 4.6 (max) on 2026-02-17, at 50.0.

Reading these numbers

The 16 scores on this page come from 2 sources: Epoch AI Benchmarking Hub (15), Scale Labs MCP Atlas leaderboard (1). 1 of them come from an official leaderboard and 15 from datasets or third-party runs. The leaderboard also lists 10 other reasoning-effort settings of this model, which redirect here. This page shows the highest.

Claude Opus 4.6 (max) is also listed at 10 other reasoning-effort settings; this page shows the highest-effort row, which is the one the leaderboard keeps by default.

Models ranked nearby

• #60 Gemini 3 Deep Think (52.6)
• #61 Claude Code + Fable 5.1 (52.6)
• #62 claude-opus-4-20250514 (52.5)
• #64 Step-3.5-Flash (52.5)
• #65 Tencent Hy4 preview (unknown) (52.4)
• #66 Muse Spark 1.1 (xhigh) (52.4)

More from Anthropic

• Claude Opus 5.5 (max) (#1, 64.3)
• Claude Fable 5.1 (max) (#3, 61.9)
• Claude Sonnet 5.5 (max) (#4, 60.8)
• Claude Opus 5 (max) (#8, 60.0)
• Claude Fable 5 (max) (#9, 59.9)
• claude-opus-4-8-xhigh-effort (#14, 56.6)

Compare it on the live leaderboard → · How the aggregate score works