Claude Sonnet 4.6
| Released | 2026-02-17 |
|---|---|
| Price per 1M | $3.00 in · $15.00 out |
| Overall score | 56.3 · rank 63 |
| Context window | 1,000,000 tokens |
| Provider | Anthropic |
| Weights | Proprietary |
Scores
| Dimension | Score | Rank | Benchmarks |
|---|---|---|---|
| Overall: Balanced | 56.3 | 63 | 38 |
| Overall: Coding | 55.1 | 69 | 23 |
| Overall: Reasoning | 56.6 | 63 | 14 |
| Overall: Agentic work | 56.5 | 66 | 21 |
| Overall: Frontend | 53.9 | 68 | 10 |
| Coding | 54.5 | 79 | 6 |
| Agents | 56.9 | 61 | 8 |
| Reasoning | 58.1 | 63 | 7 |
| Math | 55.5 | 66 | 4 |
| Knowledge | 55.1 | 58 | 3 |
| Frontend | 53.2 | 62 | 2 |
| General | 60.3 | 29 | 6 |
| Multimodal | 56.7 | 37 | 2 |
| Value vs price | +3.2 | 94 |
Configurations
| Configuration | Balanced | Benchmarks |
|---|---|---|
| Unspecified setting | 55.2 ±2.9 | 12 |
| High | 55.6 ±1.2 | 11 |
| Low | 53.1 ±7.1 | 2 |
| Max | 55.2 ±1.2 | 21 |
| Medium | 54.6 ±3.9 | 4 |
| Non-reasoning | 54.4 ±5.4 | 2 |
| Thinking | 57.9 ±4.5 | 6 |
Evidence
Coding 54.5 ±1.3
| Benchmark | Result | Source | Date |
|---|---|---|---|
| FrontierCode | 24.3% | Epoch AI | 2026-11-11 |
| DeepSWE | 29.9% | Epoch AI | 2026-11-11 |
| SciCode | 46.8% | Epoch AI | 2026-11-11 |
| LiveCodeBench (Vals run) | 82.1% | Vals AI | 2026-09-01 |
| SWE-bench Verified (Epoch run) | 75.2% | Epoch AI | 2026-02-21 |
| ProgramBench | 0.5% | Vals AI | 2026-11-09 |
| SWE-bench Verified (Vals run) | 77.4% | Vals AI | 2026-09-01 |
Agents 56.9 ±1.5
| Benchmark | Result | Source | Date |
|---|---|---|---|
| OSWorld 2.0 | 9.3% | Epoch AI | 2026-11-11 |
| APEX-Agents | 43.0% | Epoch AI | 2026-11-11 |
| Terminal-Bench 2.0 | 53.4% | Epoch AI | 2026-02-17 |
| Tax Agent Bench | 25.8% | Vals AI | 2026-11-09 |
| Terminal-Bench 2.1 (Vals run) | 57.3% | Vals AI | 2026-10-01 |
| OSWorld-Verified | 72.1% | OSWorld-Verified | 2026-03-08 |
| SkillsBench | 49.1% | Vals AI | 2026-10-29 |
| DeepResearch Bench | 54.9% | Epoch AI | 2026-11-11 |
| OSWorld (v1) | 72.1% | Epoch AI | 2026-11-11 |
| Vending-Bench 2 | $7,204 | Epoch AI | 2026-11-11 |
Reasoning 58.1 ±2.0
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Extended NYT Connections | 80.9% | Lech Mazur | 2026-11-10 |
| ARC-AGI-2 | 60.4% | Epoch AI | 2026-11-11 |
| ARC-AGI-2 (ARC Prize verified) | 60.4% | ARC Prize | 2026-11-11 |
| CritPt | 3.1% | Epoch AI | 2026-11-11 |
| ARC-AGI-1 (ARC Prize verified) | 86.5% | ARC Prize | 2026-11-11 |
| ARC-AGI-1 | 86.5% | Epoch AI | 2026-11-11 |
| GPQA Diamond | 87.4% | Epoch AI | 2026-02-20 |
| GPQA Diamond (Vals run) | 85.6% | Vals AI | 2026-09-01 |
| Chess Puzzles | 23.0% | Epoch AI | 2026-10-13 |
| Mystery Game Puzzles | 16.0% | Epoch AI | 2026-08-06 |
Math 55.5 ±1.7
| Benchmark | Result | Source | Date |
|---|---|---|---|
| FrontierMath Tiers 1-3 (2025 set) | 32.4% | Epoch AI | 2026-02-20 |
| ProofBench | 45.0% | Epoch AI | 2026-11-11 |
| FrontierMath Tier 4 (2025 set) | 8.3% | Epoch AI | 2026-02-21 |
| OTIS Mock AIME 2024-2025 | 85.8% | Epoch AI | 2026-02-20 |
Knowledge 55.1 ±3.6
| Benchmark | Result | Source | Date |
|---|---|---|---|
| SimpleQA Verified | 35.5% | Epoch AI | 2026-08-10 |
| MMLU-Pro (Vals run) | 87.3% | Vals AI | 2026-09-01 |
| Legal Research Bench | 38.5% | Vals AI | 2026-11-09 |
Frontend 53.2 ±0.7
| Benchmark | Result | Source | Date |
|---|---|---|---|
| Vibe Code Bench v1.1 | 55.8% | Vals AI | 2026-11-09 |
| LMArena WebDev | 1,522 | LMArena | 2026-11-09 |
General 60.3 ±1.6
| Benchmark | Result | Source | Date |
|---|---|---|---|
| LLM Debate Benchmark | 1,553 | Lech Mazur | 2026-10-26 |
| EQ-Bench 4 | 1,197 | EQ-Bench | 2026-10-24 |
| EQ-Bench Creative Writing v3 | 1,811 | EQ-Bench | 2026-11-11 |
| LMArena Text | 1,472 | LMArena | 2026-11-09 |
| Creative Story-Writing (Lech Mazur) | 73.0% | Lech Mazur | 2026-10-23 |
| EQ-Bench Longform Creative Writing | 79.9% | EQ-Bench | 2026-11-11 |
Multimodal 56.7 ±2.3
| Benchmark | Result | Source | Date |
|---|---|---|---|
| MMMU-Pro (Vals run) | 83.6% | Vals AI | 2026-09-01 |
| LMArena Vision | 1,275 | LMArena | 2026-11-09 |
Results not used in the score
| Benchmark | Result | Source | Why not used |
|---|---|---|---|
| ARC-AGI-1 | 86.0% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-1 (ARC Prize verified) | 86.0% | ARC Prize | A stronger configuration result was used |
| ARC-AGI-2 | 58.3% | Epoch AI | A stronger configuration result was used |
| ARC-AGI-2 (ARC Prize verified) | 58.3% | ARC Prize | A stronger configuration result was used |
| Chess Puzzles | 19.0% | Epoch AI | A stronger configuration result was used |
| Chess Puzzles | 16.0% | Epoch AI | A stronger configuration result was used |
| DeepResearch Bench | 50.4% | Epoch AI | A stronger configuration result was used |
| Extended NYT Connections | 76.4% | Lech Mazur | A stronger configuration result was used |
| Extended NYT Connections | 44.8% | Lech Mazur | A stronger configuration result was used |
| GPQA Diamond | 78.8% | Epoch AI | A stronger configuration result was used |
| GPQA Diamond | 83.3% | Epoch AI | A stronger configuration result was used |
| GPQA Diamond | 83.3% | Epoch AI | A stronger configuration result was used |
| LLM Debate Benchmark | 1,534 | Lech Mazur | A stronger configuration result was used |
| Mystery Game Puzzles | 14.0% | Epoch AI | A stronger configuration result was used |
| OSWorld 2.0 | 8.3% | Epoch AI | A stronger configuration result was used |
| OTIS Mock AIME 2024-2025 | 71.1% | Epoch AI | A stronger configuration result was used |
| OTIS Mock AIME 2024-2025 | 75.6% | Epoch AI | A stronger configuration result was used |
| OTIS Mock AIME 2024-2025 | 82.2% | Epoch AI | A stronger configuration result was used |
| SimpleQA Verified | 32.8% | Epoch AI | A stronger configuration result was used |
| Vibe Code Bench v1.1 | 51.5% | Vals AI | A stronger configuration result was used |
Names matched to this model
claude-sonnet-4-6