Claude Sonnet 4.6 (max): benchmark scores, price and ranking
Claude Sonnet 4.6 (max) from Anthropic ranks #217 of 743 LLMs on the LLMs Tiger aggregate benchmark score (50.0), based on 12 benchmark results captured up to 2026-10-13.
- Organisation
- Anthropic
- Type
- Frontier model
- Open weights
- No
- Released
- 2026-02-17 (source: epoch)
- Cheapest API price
- $3.00 in / $15.00 out per million tokens (via anthropic, observed 2026-10-13T19:47Z)
- Aggregate rank
- #217 of 743 (score 50.0)
Relative to other models, Claude Sonnet 4.6 (max) is strongest on LMCA (#18 of 108) and weakest on Chess Puzzles (#88 of 122).
Across the 11 benchmarks where at least ten models are scored, Claude Sonnet 4.6 (max) is in the top 10% on 0 and in the top half on 8.
Its cheapest listed input price, $3.00 per million tokens, is lower than that of 11% of the 427 models with a listed price.
Claude Sonnet 4.6 (max) benchmark results
| Benchmark | Score | Rank | Source | Captured |
|---|---|---|---|---|
| LMCA | 46.5% | #18 of 108 | Epoch AI Benchmarking Hub | 2026-10-13 |
| DTBench | 89.9% | #30 of 146 | Epoch AI Benchmarking Hub | 2026-10-13 |
| ProofBench | 45.0% | #18 of 46 | Epoch AI Benchmarking Hub | 2026-10-13 |
| OTIS Mock AIME | 71.1% | #75 of 189 | Epoch AI Benchmarking Hub | 2026-10-13 |
| ARC-AGI-2 | 58.3% | #35 of 87 | ARC Prize official leaderboard | 2026-10-13 |
| GPQA Diamond | 78.8% | #112 of 275 | Epoch AI Benchmarking Hub | 2026-10-13 |
| SciCode | 46.8% | #48 of 106 | Epoch AI Benchmarking Hub | 2026-10-13 |
| CritPt | 3.1% | #54 of 115 | Epoch AI Benchmarking Hub | 2026-10-13 |
| OSWorld 2 | 8.3% | #5 of 8 | Epoch AI Benchmarking Hub | 2026-10-13 |
| GDP.pdf | 18.0% | #17 of 25 | Epoch AI Benchmarking Hub | 2026-10-13 |
| SimpleQA Verified | 32.8% | #47 of 69 | Epoch AI Benchmarking Hub | 2026-10-13 |
| Chess Puzzles | 3.0% | #88 of 122 | Epoch AI Benchmarking Hub | 2026-10-13 |
About Claude Sonnet 4.6 (max)
Claude Sonnet 4.6 (max) is a frontier model from Anthropic. It was released on 2026-02-17, according to Epoch AI. The cheapest API listing we track charges $3.00 per million input tokens and $15.00 per million output tokens, through anthropic. We hold 12 benchmark results for it across 4 areas, and none is in the top tenth of its board.
Where Claude Sonnet 4.6 (max) stands
On our aggregate score, Claude Sonnet 4.6 (max) is #217 of 743 ranked models, around the middle of the table. Its score is 50.0; the scale is centred on 50, so it sits 0.0 points above the typical model. That beats 71% of the ranked field. Among the 70 ranked Anthropic models it is #36. The next one up is CodeAct v1.8 at 50.3. The next one down is claude-sonnet-4-20250514_2K at 50.0.
Claude Sonnet 4.6 (max), benchmark by benchmark
LMCA puts Claude Sonnet 4.6 (max) at 46.5%, #18 of 108. That beats 83% of the models tested, but 21.7 points short of Claude Opus 5.5 (max) (68.2%). The results either side are Muse Spark 1.2 (xhigh) (48.4%) and Qwen3.8 Max (xhigh) (46.2%). It covers language-model capability assessment. The board median is 29.2%, so Claude Sonnet 4.6 (max) is 17.3 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
On DTBench, Claude Sonnet 4.6 (max) reaches 89.9%. It ranks #30 of 146, ahead of 79% of the field, and 9.1 points behind Claude Opus 5.5 (max) at 98.9%. The results either side are nemotron-3-ultra (90.1%) and GPT-5.6 Luna (max) (88.8%). It covers dynamic tool-use benchmark. The board median is 68.8%, so Claude Sonnet 4.6 (max) is 21.1 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Claude Sonnet 4.6 (max) manages 45.0% on ProofBench, good for #18 of 46. Claude Fable 5.1 (max) leads at 100.0%, 55.0 points further up, and 61% of models sit below. The results either side are GLM-5.3 (max) (49.0%) and Grok 4.5 (high) (31.0%). It covers writing mathematical proofs. The board median is 21.0%, so Claude Sonnet 4.6 (max) is 24.0 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
OTIS Mock AIME puts Claude Sonnet 4.6 (max) at 71.1%, #75 of 189. That beats 60% of the models tested, but 28.9 points short of Claude Fable 5.1 (max) (100.0%). The results either side are Gemini 2.5 Flash Preview (Apr 2025) (73.1%) and claude-sonnet-4-20250514_32K (71.1%). It covers competition maths (AIME 2024-25 style). The board median is 57.8%, so Claude Sonnet 4.6 (max) is 13.3 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
On ARC-AGI-2, Claude Sonnet 4.6 (max) reaches 58.3%. It ranks #35 of 87, ahead of 60% of the field, and 36.7 points behind GPT-6 Astra (max) at 95.0%. The results either side are GPT-6 Luna (max) (59.3%) and GPT-5.2 (xhigh) (52.9%). It covers abstract reasoning, ARC Prize public eval set. The board median is 22.8%, so Claude Sonnet 4.6 (max) is 35.5 points above it. Source: ARC Prize official leaderboard, captured 2026-10-13.
Claude Sonnet 4.6 (max) manages 78.8% on GPQA Diamond, good for #112 of 275. GPT-6 Astra (max) leads at 95.8%, 17.0 points further up, and 59% of models sit below. The results either side are Qwen3.5 9B (79.0%) and DeepSeek R1 (0528) (78.8%). It covers expert-level science questions. The board median is 71.7%, so Claude Sonnet 4.6 (max) is 7.1 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
CritPt puts Claude Sonnet 4.6 (max) at 3.1%, #54 of 115. That beats 53% of the models tested, but 29.1 points short of GPT-5.6 Sol (max) (32.3%). The results either side are Ring-2.6-1T (3.7%) and Kimi K2.5 (3.1%). It covers unpublished research-level physics problems. The board median is 2.9%, so Claude Sonnet 4.6 (max) is 0.3 points above it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Claude Sonnet 4.6 (max) lands below the middle on GDP.pdf: 18.0%, #17 of 25, with 64% of the field scoring higher. GPT-6 Astra (max) leads at 34.2%; the closest result above is Kimi K3 (max) at 19.0%. It covers economically valuable document tasks. The board median is 22.8%, so Claude Sonnet 4.6 (max) is 4.8 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Claude Sonnet 4.6 (max) lands below the middle on SimpleQA Verified: 32.8%, #47 of 69, with 67% of the field scoring higher. gemini-4-argon-high leads at 76.0%; the closest result above is grok-4.3 (high) at 33.2%. It covers short factual questions. The board median is 41.0%, so Claude Sonnet 4.6 (max) is 8.2 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
Claude Sonnet 4.6 (max) lands below the middle on Chess Puzzles: 3.0%, #88 of 122, with 71% of the field scoring higher. GPT-6 Astra (max) leads at 72.0%; the closest result above is Qwen3-Max-Instruct at 4.0%. It covers chess tactics puzzles. The board median is 12.5%, so Claude Sonnet 4.6 (max) is 9.5 points below it. Source: Epoch AI Benchmarking Hub, captured 2026-10-13.
The full table above also lists the other 2 results, including boards with fewer than ten models, where a rank says little.
Strengths and weaknesses
Reasoning, maths and science: solid, ahead of 59% of the field on average across 5 benchmarks. Its best board here is ProofBench (#18 of 46) and its weakest is CritPt (#54 of 115). Agents and tool use: solid, ahead of 56% of the field on average across 2 benchmarks. Its best board here is DTBench (#30 of 146) and its weakest is GDP.pdf (#17 of 25). General knowledge and chat: uneven, ahead of 48% of the field on average across 3 benchmarks. Its best board here is LMCA (#18 of 108) and its weakest is Chess Puzzles (#88 of 122).
Price and value
The median input price across the 427 priced models is $0.500 per million tokens, so Claude Sonnet 4.6 (max) costs more than the typical listing. Models that score within 1.5 points or better and cost less: Qwen2.5-Coder-7B-Instruct ($0.010, 49.5), DeepSeek V4.1 Flash (high) ($0.010, 48.9), Qwen3.7 Flash ($0.030, 50.5). Similar scores cost more on Muse Wonder (max) ($4.00) and Claude Opus 4.6 (120K, High) ($5.00).
Release history
The Anthropic model released before it was Claude Opus 4.6 (max) on 2026-02-05, which scores 52.5. The next one was Claude Opus 4.7 (max) on 2026-04-16, at 53.7.
Reading these numbers
The 12 scores on this page come from 2 sources: Epoch AI Benchmarking Hub (11), ARC Prize official leaderboard (1). 1 of them come from an official leaderboard and 11 from datasets or third-party runs. The leaderboard also lists 8 other reasoning-effort settings of this model, which redirect here. This page shows the highest.
Claude Sonnet 4.6 (max) is also listed at 8 other reasoning-effort settings; this page shows the highest-effort row, which is the one the leaderboard keeps by default.
Models ranked nearby
- #214 Gemini 3 Deep Think (Preview) ² (50.1)
- #215 glm-4.7-flash (50.1)
- #216 grok-3-beta (50.0)
- #218 grok-4-20 (50.0)
- #219 step-5-preview-high (50.0)
- #220 claude-sonnet-4-20250514_2K (50.0)
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