Automated page. Written and translated automatically by AI and checked against the sources by software; no human editor. SI-IQ is an index built from public benchmark results, not an IQ test. How it works
How smart is Claude Sonnet 4.5?
Claude Sonnet 4.5 is a model from Anthropic, released on Sep 29, 2025. Its SI-IQ is 110 (90% interval 107–113), estimated from 11 benchmark results by 3 independent evaluators. Its strongest area is Agents and its weakest is Reasoning.
Facts
- Developer
- Anthropic
- Released
- Sep 29, 2025
- Access
- API access
- Benchmarks used
- 11 results from 3 evaluators
- API price
- $3.00 input / $15 output per 1M tokens (developer's price list, Oct 18, 2026)
- Context window
- 200,000 tokens
Ability by area (SI-IQ, 90% interval)
Every benchmark result we used
Each score is one value exactly as the source published it: for Epoch AI the best of its run settings, for OpenRouter the median hosting provider (the developer's own hosting when it ran the test). “Implied SI-IQ” is the ability that this single result points to; the overall SI-IQ is their precision-weighted average.
| Benchmark | Score | Evaluated by | Source |
|---|---|---|---|
| Chess Puzzles | 10.0% | Epoch AI | Epoch AI Benchmarking Hub · data file |
| Mystery Game Puzzles | 17.0% | Epoch AI | Epoch AI Benchmarking Hub · data file |
| FrontierMath Tier 4 | 2.4% | Epoch AI | Epoch AI Benchmarking Hub · data file |
| FrontierMath Tiers 1–3 | 23.9% | Epoch AI | Epoch AI Benchmarking Hub · data file |
| MATH Level 5 | 97.7% | Epoch AI | Epoch AI Benchmarking Hub · data file |
| OTIS Mock AIME 2024–2025 | 77.8% | Epoch AI | Epoch AI Benchmarking Hub · data file |
| SWE-bench Verified (Epoch AI run) | 71.3% | Epoch AI | Epoch AI Benchmarking Hub · data file |
| GPQA Diamond | 82.3% | Epoch AI | Epoch AI Benchmarking Hub · data file |
| GPQA Diamond (OpenRouter run) | 81.1% | OpenRouter | OpenRouter benchmark datasets · data file |
| SimpleQA Verified | 30.7% | Epoch AI | Epoch AI Benchmarking Hub · data file |
| Berkeley Function Calling Leaderboard | 73.2% | UC Berkeley (Gorilla) | Berkeley Function Calling Leaderboard · data file |
| τ²-bench Airline (OpenRouter run) | 75.3% | OpenRouter | OpenRouter benchmark datasets · data file |
Human preference (Arena)
On the Arena leaderboard, where people vote between two anonymous answers, Claude Sonnet 4.5 has a rating of 1,455 (1,452–1,458) from 82,688 votes (published Oct 15, 2026). Human preference is a different measure from benchmark ability and is not part of SI-IQ. Arena (Hugging Face)
Models with similar SI-IQ
| Model | SI-IQ | |
|---|---|---|
| Gemini 3.5 Flash-LiteGoogle DeepMind | 109 | Claude Sonnet 4.5 vs Gemini 3.5 Flash-Lite |
| Qwen 3.6 35B-A3BAlibaba | 111 | Claude Sonnet 4.5 vs Qwen 3.6 35B-A3B |
| Gemini 2.5 Pro (Jun 2025)Google DeepMind | 108 | Claude Sonnet 4.5 vs Gemini 2.5 Pro (Jun 2025) |
| Qwen3-235B-A22B-Thinking (Jul 2025)Alibaba | 111 | Claude Sonnet 4.5 vs Qwen3-235B-A22B-Thinking (Jul 2025) |
| MiniMax-M3MiniMax | 108 | Claude Sonnet 4.5 vs MiniMax-M3 |
Computed with SI-IQ method version 1.0.0. Read the full methodology
Data sources: Arena leaderboard dataset on Hugging Face (CC BY 4.0); Berkeley Function Calling Leaderboard, Gorilla project (Apache-2.0); Epoch AI, ‘Capabilities & benchmarking’ (CC BY 4.0); API list prices via LiteLLM's price table (MIT); each price links to the developer's pricing page; Source: OpenRouter (openrouter.ai), licensed under CC BY 4.0: reuse and republish with attribution to OpenRouter..