Anthropic model product
Claude Sonnet 4.6
6 is a full upgrade of the model's skills across coding, computer use, long-context reasoning, agent planning, knowledge work, and design.
Updated Aug 12, 2026. Default version: Claude Sonnet 4.6
LLMBoard score72.6Claude Sonnet 4.6
Coverage60%17 benchmark families
Context window1MTokens
Official input price$3Anthropic API
Claude Sonnet 4.6 Specifications
Technical details for the model's default version.
- Version
- Claude Sonnet 4.6
- Released
- Feb 17, 2026
- Knowledge cutoff
- Aug 31, 2025
- Parameters
- N/A
- Context window
- 1M
- Max output
- 64K
- Inputs
- audio, image, pdf, text, video
- Outputs
- text
- Open weights
- No
- License
- Proprietary
Claude Sonnet 4.6 Capability Profile
This profile uses the model's current scored version. Arena ratings and prices are shown separately.
Claude Sonnet 4.6 Benchmark Results
Benchmark scores for Claude Sonnet 4.6.
| Finance Agent | 63.3% | 02 | 8 | 85.7% | C | |
| Tau2 Retail | 91.7% | 02 | 26 | 96.0% | C | |
| OSWorld | 72.5% | 04 | 20 | 84.2% | C | |
| Legal Agent Benchmark | 5.4% | 05 | 13 | 66.7% | B | |
| ARC-AGI v2 | 58.3% | 06 | 16 | 66.7% | C | |
| Finance Agent v2 | 51.0% | 07 | 26 | 76.0% | B | |
| Tau2 Telecom | 97.9% | 07 | 35 | 82.3% | C | |
| MMMLU | 89.3% | 12 | 49 | 77.1% | C | |
| LiveBench | 75.5% | 13 | 38 | 67.6% | B | |
| SWE-Bench Verified | 79.6% | 15 | 105 | 86.5% | C | |
| DeepSWE 1.1 | 30.0% | 19 | 20 | 5.3% | B | |
| Humanity's Last Exam | 49.0% | 21 | 93 | 78.3% | C | |
| Terminal-Bench 2.0 | 59.1% | 23 | 49 | 54.2% | C | |
| GPQA | 89.9% | 25 | 234 | 89.7% | C | |
| MCP Atlas | 61.3% | 26 | 31 | 16.7% | C | |
| BrowseComp | 74.7% | 27 | 58 | 54.4% | C | |
| MMMU-Pro | 75.6% | 31 | 66 | 53.9% | C |
Claude Sonnet 4.6 Arena Results
Preference and agent-evaluation results for the default version.
| search | overall | 06 | 1220.8 | 111,959 | N/A | |
| document | overall | 08 | 1482.8 | 54,427 | N/A | |
| search style control | overall | 08 | 1199.6 | 111,959 | N/A | |
| search factuality | overall | 09 | 1201.0 | 111,907 | N/A | |
| agent bash recovery steps | overall | 12 | 0.1 | N/A | 24.2K | |
| document style control | overall | 14 | 1477.5 | 54,427 | N/A | |
| agent | overall | 20 | 0.0 | N/A | 1.2M | |
| agent steerability | overall | 20 | 0.0 | N/A | 36.4K | |
| webdev | overall | 21 | 1524.1 | 19,163 | N/A | |
| agent tool hallucination | overall | 22 | 0.0 | N/A | 1.1M | |
| agent praise complaint | overall | 24 | 0.0 | N/A | 9.4K | |
| vision | overall | 25 | 1282.2 | 24,997 | N/A | |
| agent task outcome explicit | overall | 26 | 0.0 | N/A | 26.9K | |
| vision style control | overall | 26 | 1275.3 | 24,997 | N/A | |
| text | overall | 29 | 1457.7 | 66,159 | N/A | |
| text style control | overall | 30 | 1472.2 | 66,159 | N/A | |
| text factuality | overall | 32 | 1460.1 | 65,872 | N/A |
Claude Sonnet 4.6 Runtime Performance
Provider-specific output speed and catalog latency for Claude Sonnet 4.6. Runtime does not affect the capability score.
| Anthropic | 2.808 tok/s | 2.646 s | 200K | 64K |
Output Speed is generated output tokens received per second. Catalog latency is reported separately from observed provider TTFT.
Claude Sonnet 4.6 Pricing
Official vendor API pricing appears first, followed by individual provider offers.
- Official API
- $3 input, $15 output per 1M
- Official provider
- Anthropic
- Lowest third-party
- From $0.90 input, $5.55 output per 1M via Xpersona
- Tracked offerings
- 42
| Xpersona | claude-sonnet-4-6 | global | $0.90 | $5.55 | 200K | |
| Poe | anthropic/claude-sonnet-4.6 | global | $2.6 | $13 | 983K | |
| Impossibl | anthropic/claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| CrossModel | anthropic/claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| NanoGPT | anthropic/claude-sonnet-4.6 | global | $3 | $15 | 1M | |
| FastRouter | anthropic/claude-sonnet-4.6 | global | $3 | $15 | 1M | |
| NEAR AI Cloud | anthropic/claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| DaoXE | claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| LLM Gateway | claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| OpenCode Zen | claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| GitHub Copilot | claude-sonnet-4.6 | global | $3 | $15 | 200K | |
| 302.AI | claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| Neon | claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| FrogBot | claude-sonnet-4-6 | global | $3 | $15 | 200K | |
| Pioneer | claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| ZenMux | anthropic/claude-sonnet-4.6 | global | $3 | $15 | 1M | |
| Abacus | claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| Auriko | claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| Azure | claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| Vertex (Anthropic) | claude-sonnet-4-6@default | global | $3 | $15 | 1M | |
| AIHubMix | claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| Ofox | anthropic/claude-sonnet-4.6 | global | $3 | $15 | 1M | |
| OrcaRouter | anthropic/claude-sonnet-4.6 | global | $3 | $15 | 1M | |
| routing.run | claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| Vertex | claude-sonnet-4-6@default | global | $3 | $15 | 1M | |
| GMI Cloud | anthropic/claude-sonnet-4.6 | global | $3 | $15 | 409.6K | |
| FreeModel | claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| Amazon Bedrock | anthropic.claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| Requesty | claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| Kilo Gateway | anthropic/claude-sonnet-4.6 | global | $3 | $15 | 1M | |
| Merge Gateway | anthropic/claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| Modelis | claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| Vercel AI Gateway | anthropic/claude-sonnet-4.6 | global | $3 | $15 | 1M | |
| OpenRouter | anthropic/claude-sonnet-4.6 | global | $3 | $15 | 1M | |
| Anthropic | claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| Cloudflare AI Gateway | anthropic/claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| Perplexity Agent | anthropic/claude-sonnet-4-6 | global | $3 | $15 | 200K | |
| Databricks | databricks-claude-sonnet-4-6 | global | $3 | $15 | 1M | |
| Cortecs | claude-4-6-sonnet | global | $3.2 | $15.94 | 1M | |
| Venice AI | claude-sonnet-4-6 | global | $3.6 | $18 | 1M |
Official prices use only the vendor's configured official Provider and positive standard USD PAYG rates. Third-party offers remain explicitly labeled.
Claude Sonnet 4.6 Versions
Available versions of this model. The score column identifies the version used in the overall ranking.
| Claude Sonnet 4.6 | 72.6 | N/A | 1M | 64K | No | Proprietary |
What is Claude Sonnet 4.6?
Key information about Claude Sonnet 4.6 and its available data.
6 is a full upgrade of the model's skills across coding, computer use, long-context reasoning, agent planning, knowledge work, and design. 5 approximately 70% of the time. First Sonnet-class model with 1M token context window (beta) and context compaction.
Major improvement in computer use skills compared to prior Sonnet models. Default model on Free and Pro plans. Pricing: $3/$15 per million tokens (input/output).
Data as of 2026-08-11.
Claude Sonnet 4.6 vs nearby models
Open a comparison with the three ranked models immediately above and below this model.
Models similar to Claude Sonnet 4.6
Recommendations prioritize the same model type and family, then the closest LLMBoard score.
AN
Claude Opus 4.5
Anthropic
AN
Claude Opus 4.6
Anthropic
AN
Claude Sonnet 5
Anthropic
AN
Claude Opus 4.7
Anthropic
AN
Claude Opus 4.8
Anthropic
AN
Claude Sonnet 4.5
Anthropic
FAQ
Common questions about Claude Sonnet 4.6.
When was Claude Sonnet 4.6 released?
Claude Sonnet 4.6's default version was released on Feb 17, 2026.
How much does Claude Sonnet 4.6 cost?
Claude Sonnet 4.6's official API price is $3 per million input tokens and $15 per million output tokens via Anthropic. The lowest tracked third-party offer starts at $0.90 input and $5.55 output via Xpersona.
Who created Claude Sonnet 4.6?
Claude Sonnet 4.6 was created by Anthropic.
What is the context window for Claude Sonnet 4.6?
The default version has a 1M token context window.
Is Claude Sonnet 4.6 open weight?
No. The default version is not marked as having publicly available weights.
How many API providers offer Claude Sonnet 4.6?
42 provider offerings are linked to the default version.
What models should I compare Claude Sonnet 4.6 with?
Nearby ranked alternatives include Qwen3.7 Plus, Muse Spark, Seed 2.1 Turbo.