## Wednesday, February 18, 2026

# Claude Sonnet 4.6 for RAG

Umida Muratbekova

RAGLlm Evals

Anthropic released [Claude Sonnet 4.6](/content/llms/claude-sonnet-46/index.html), and we added it to our [LLM-for-RAG leaderboard](/content/llms/index.html) to test three workloads:

- Factual QA (web-style passages)
- Long-form synthesis (essay / narrative content)
- Scientific verification (claims grounded in papers)

**TL;DR**: Sonnet 4.6 placed #3 overall, behind Opus 4.6 and GPT-5.1. Biggest win was long-form synthesis. Smallest edge is on scientific verification.

## Key Findings

### Factual answers

Sonnet 4.6 was strong on factual QA. It generally pulled the answer directly from retrieved passages and avoided unnecessary filler.

It ranked #2 with a 71% win rate, well ahead of [Sonnet 4.5](/content/llms/claude-sonnet-45/index.html) at 59%. In practice, this shows up as cleaner, more grounded answers.

### Synthesis: biggest improvement

This is where Sonnet 4.6 separates from 4.5.

On long-form synthesis, [Sonnet 4.6](/content/llms/claude-sonnet-46/index.html) wins nearly 2x more often than [4.5](/content/llms/claude-sonnet-45/index.html) (53% vs 27%). This is the largest gap across all three workloads.

The extra quality comes with extra depth. [Sonnet 4.6](/content/llms/claude-sonnet-46/index.html) averaged 498 tokens per answer vs 336 for [Sonnet 4.5](/content/llms/claude-sonnet-45/index.html) (+48%).

In practice, 4.6 connects ideas across chunks instead of just listing what each source says.

### Scientific claims: **smallest gap**

Sonnet 4.6 still leads on scientific tasks, but it's narrow:

[Sonnet 4.6](/content/llms/claude-sonnet-46/index.html) edges out [4.5](/content/llms/claude-sonnet-45/index.html) by just 5 points (50% vs 45% win rate) - the smallest margin across all workloads.

The pattern fits: [4.5](/content/llms/claude-sonnet-45/index.html)'s literal, conservative style is more competitive on strict verification, while [4.6](/content/llms/claude-sonnet-46/index.html)'s explanatory depth has less advantage here.

## Takeaway

[Sonnet 4.6](/content/llms/claude-sonnet-46/index.html) is a **strong general-purpose RAG model** (#3 overall). Its biggest gain is long-form synthesis, where it nearly doubles [4.5](/content/llms/claude-sonnet-45/index.html)'s win rate.

For best overall quality, use [Opus 4.6](/content/llms/claude-opus-46/index.html) or [GPT-5.1](/content/llms/gpt-51/index.html) (top 2 on the leaderboard). If your workload is mostly factual Q&A and synthesis and you want a lighter option, [Sonnet 4.6](/content/llms/claude-sonnet-46/index.html) is the better pick. For scientific verification, 4.6 and 4.5 are close: [4.6](/content/llms/claude-sonnet-46/index.html) explains more, [4.5](/content/llms/claude-sonnet-45/index.html) is shorter and more literal.

See the full rankings and per-task breakdown on the [LLM-for-RAG leaderboard](/content/llms/index.html).