Overview
The original Claude 3.5 Sonnet was released on June 20, 2024, and was a breakout model: it outperformed the much larger Claude 3 Opus on most benchmarks while keeping mid-tier pricing and faster speed. It set new marks for the time on graduate-level reasoning (GPQA), undergraduate knowledge (MMLU), and coding (HumanEval).
On the well-documented benchmarks it scored 59.4% on GPQA Diamond, 88.7% on MMLU, 92.0% on HumanEval, and 68.3% on MMMU for vision. In Anthropic's internal agentic coding evaluation it solved 64% of problems, versus 38% for Claude 3 Opus. The launch also introduced Artifacts on claude.ai, a workspace for viewing and iterating on generated content.
Claude 3.5 Sonnet has a 200K token context window, up to 8,192 output tokens, and supports text and vision input, at $3 / $15 per million tokens. Its model ID is claude-3-5-sonnet-20240620; the snapshot was retired on October 28, 2025, with Sonnet 4.6 as the recommended successor.
| Released | 2024-06-20 |
|---|---|
| License | Proprietary |
| Weights | API only |
| Context | 200K |
| Max output | 8K |
| Architecture | Proprietary transformer (non-reasoning); Claude 3.5 generation. |
| Knowledge cutoff | April 2024 |
| Modalities | Text, Vision, PDF |
| Status | Deprecated |
Benchmarks
- GPQA Diamond (0-shot CoT)59.4%
- MMLU (5-shot)88.7%
- HumanEval (0-shot)92%
- MMMU (vision)68.3%
Scores on a 0–100 scale (25-point gridlines); higher is better. Each benchmark links to its published source.
Pricing
| Input | $3.00 per million tokens |
|---|---|
| Output | $15.00 per million tokens |
One-fifth the cost of Claude 3 Opus at launch.
Strengths
- Outperformed the larger Claude 3 Opus on most benchmarks at one-fifth the cost
- Strong coding (92.0% HumanEval) and reasoning (59.4% GPQA Diamond)
- Best-in-class vision at launch (68.3% MMMU)
- Introduced Artifacts for interactive, iterative output on claude.ai
- Roughly 2x the speed of Claude 3 Opus
Best for
- General coding and code generation
- Graduate- and undergraduate-level reasoning and knowledge work
- Vision tasks: charts, diagrams and document understanding
- Interactive content generation via Artifacts
- Cost-efficient replacement for larger flagship models
How to access
| Provider | Model ID |
|---|---|
| Anthropic ↗ | claude-3-5-sonnet-20240620 |
| Amazon Bedrock ↗ | anthropic.claude-3-5-sonnet-20240620-v1:0 |
Claude Sonnet — every version
The full lineage of the Claude Sonnet line, newest first. Every version has its own page — click any to compare specs, benchmarks and pricing.
| Version | Released | Context | License |
|---|---|---|---|
| Claude Sonnet 4.6current | 2026-02-17 | 1M | Proprietary |
| Claude Sonnet 4.5 | 2025-09-29 | 1M | Proprietary |
| Claude Sonnet 4 | 2025-05-22 | 200K | Proprietary |
| Claude 3.7 Sonnet | 2025-02-24 | 200K | Proprietary |
| Claude 3.5 Sonnet (new) | 2024-10-22 | 200K | Proprietary |
| Claude 3.5 Sonnet | 2024-06-20 | 200K | Proprietary |
| Claude 3 Sonnet | 2024-03-04 | 200K | Proprietary |
FAQ
Why was the original Claude 3.5 Sonnet significant?
Released June 20, 2024, it beat the larger Claude 3 Opus on most benchmarks while keeping mid-tier price and faster speed, and it introduced Artifacts on claude.ai.
What were its benchmark scores?
59.4% on GPQA Diamond, 88.7% on MMLU, 92.0% on HumanEval, and 68.3% on MMMU for vision.
What is the model ID and context window?
The model ID is claude-3-5-sonnet-20240620, with a 200K token context window and up to 8,192 output tokens.
Is it still available?
No. The June 2024 snapshot was retired on October 28, 2025. Anthropic recommends Claude Sonnet 4.6.