Anthropic · Released Feb 17, 2026Claude Sonnet 4.6
Claude workhorse for coding agents, careful analysis, and production cost control.
- Capability
- 152.3ECI · #40 of 148
- Input
- $3.00per 1M tokens
- Output
- $15.00per 1M tokens
- Context
- 1M64K max output
What it costs
- Input
- $3.00per million tokens
- Output
- $15.00per million tokens
- Cached input
- $0.30per million tokens
- Cache write
- $3.75per million tokens
- Blended (3:1)
- $6.00pricier than 90% of priced models
Typical monthly bills
| Monthly usage | Estimated cost |
|---|---|
| Side project2M input + 0.5M output tokens | $13.50 |
| Team assistant25M input + 5M output tokens | $150.00 |
| Production app250M input + 50M output tokens | $1,500 |
Official Anthropic API price, as listed on models.dev.
Independent benchmarks
#40 of 148 scored models
The shaded band is Epoch AI’s confidence range (149.6–154.7); the tick marks the median scored model.
Scores from Epoch AI, run independently of Anthropic.
The details
- Lab
- Anthropic
- Released
- Feb 17, 2026
- Knowledge cutoff
- Aug 31, 2025
- Context window
- 1,000,000 tokens
- Max output
- 64,000 tokens
- Inputs
- Text, Images, PDFs
- Output
- Text
- Reasoning
- Adjustable effort low · medium · high · max
- Tool calling
- Yes
- Structured output
- Yes
- Weights
- Proprietary
- API model ID
claude-sonnet-4-6on Anthropic- Availability
- 41 API providerslisted on models.dev
What Anthropic claims
Published by the lab at launch. Settings vary, so compare these only with care.
| Benchmark | Score | Setting | Source |
|---|---|---|---|
| SWE-Atlas Codebase QnA | 31.2 | — | Source |
| SWE-Atlas Refactoring | 32.21 | — | Source |
| SWE-Atlas Test Writing | 31.76 | — | Source |
| Artificial Analysis Coding Agent Index | 49.4 | medium | Source |
| SWE-Atlas Codebase QnA | 70.3 | medium | Source |
| SWE-Bench Pro | 14.9 | medium | Source |
| Terminal-Bench v2.1 | 63.1 | medium | Source |
| Terminal-Bench v2.1 | 67 success rate | — | Source |
| Humanity's Last Exam | 34.6 | no tools | Source |
| Humanity's Last Exam | 46.8 | with tools | Source |
| OSWorld-Verified | 78.5 success rate | — | Source |
More from Anthropic
About Claude Sonnet 4.6
How much does Claude Sonnet 4.6 cost?
Claude Sonnet 4.6 costs $3.00 input / $15.00 output per million tokens (official Anthropic API price). Cached input is $0.30 per million tokens, and writing to the cache costs $3.75. At a 3:1 input-to-output mix that is $6.00 per million tokens, more expensive than 90% of the 360 priced models we track.
What is the context window of Claude Sonnet 4.6?
Claude Sonnet 4.6 accepts up to 1,000,000 tokens per request and can write up to 64,000 tokens in one response.
How good is Claude Sonnet 4.6?
Epoch AI gives Claude Sonnet 4.6 a Capabilities Index score of 152.3 (likely range 149.6–154.7), ranking it #40 of 148 models Epoch has scored. Epoch AI benchmark results: GPQA Diamond 87.4%, OTIS Mock AIME 2024–2025 85.8%, SWE-bench Verified 75.2%, SimpleQA Verified 35.5%.
Is Claude Sonnet 4.6 open source?
No. Claude Sonnet 4.6 is proprietary; you use it through Anthropic’s API or partner platforms.
What inputs does Claude Sonnet 4.6 support?
Claude Sonnet 4.6 accepts text, images and PDFs and replies in text. It is a reasoning model with low, medium, high and max effort settings, supports tool calling and can return structured JSON output.
When was Claude Sonnet 4.6 released?
Anthropic released Claude Sonnet 4.6 on Feb 17, 2026. Its training data runs to Aug 31, 2025.
What are the best alternatives to Claude Sonnet 4.6?
The closest current models from other labs on capability, price and release date are GPT-5.2 (OpenAI, ECI 153.5, $1.75 / $14.00), Gemini 3 Pro Preview (Google, ECI 153.0, $2.00 / $12.00), Qwen3.7 Max (Alibaba (Qwen), ECI 153.7, $2.50 / $7.50) and Inkling (Thinking Machines, ECI 148.6, $3.74 / $9.36).