Claude Sonnet 4.6
anthropic/claude-sonnet-4.6- Anthropic
- Family · Claude Sonnet 4
- Released 17 Feb 2026
- Text · Image · File → Text
- Reasoning · adjustable
- Tools · JSON · Vision
Not on OSMEXO yet
Spec & list price
Context1M
Max output128K
Input / 1M$3.00
Output / 1M$15.005× input
Scores · standing
Intelligence30.1#66/489
Coding63.0#36/196
Usage
Tokens / day29.6B▼ 19% 7d
Usage rank#61
Requests / day1.7M
Tool calls / req0.18
Per request · 30 days
Avg input14.9Kabout median
Avg output5310.8× median
Cache hit60%0.9× median
Reasoning share5.2%0.1× median
Tool-call failures1.5%0.8× median
Consumption
Peak 57.2B 9 OctLow 24.1B 17 OctRank today #61
News & trends
3 mentions this week ▼ 70%this model · videos, posts, articles · 30 days
Trending · 14 days
- Claude Sonnet 4.6 JAILBREAK 🔥 Como Desbloquear o Modo SEM CENSURA e Usar o MÁXIMO da IA!3.9K views
- Claude Ai Jailbreak 🔓 || How to Jailbreak Claude sonnet 4.6 Max think2.1K views
News · latest from Anthropic
- Claude Opus 5.5 Can Draw Permit-Level Architectural Drawingsstephenjgallagher.substack.com · 2 Nov
- GPT-6.1 Terra vs. Claude Opus 5.5: Cheaper and faster in three tests at maxthenewstack.io · 2 Nov
- Claude Opus 5.5 and Claude Sonnet 5.5 are now available in Google AntigravityGoogle Cloud · 20 Oct
- GLM-5.3 vs. Claude Fable 5 on DeepSWE: Cost, Coding, and RoutingTogether AI · 21 Aug
Benchmarks
bar = share of the top score · Benchmark data: Artificial AnalysisSpeed · median across 7 providers, latest day
Output speed#193/33458 tok/s1.1× slower than median
Fastest provider123 tok/sbest of 7 providers
First token#1310.81 s1.3× quicker than median
#8 cheapest of 10 with a similar score
Value
Intelligence Index
Blended price per 1M tokens (log)
Technical details
- Reasoning
- on · medium
- Tokenizer
- Claude
- Moderation
- moderated
Reasoning effort
- max
- high
- medium · default
- low
- none
Request parameters 14
- tools
- tool_choice
- response_format
- structured_outputs
- reasoning
- reasoning_effort
- include_reasoning
- temperature
- top_p
- top_k
+4 more
- max_tokens
- max_completion_tokens
- stop
- verbosity
When to use
Best for
- Reasoning-heavy chat and analysis
- Data analysis over documents
- Tool-calling agents with structured output
- Long document question answering
Think twice for
- Long autonomous agent runs (weakest suite)
- Hardest coding agent tasks
- Self-hosting (closed weights)
Research papers
- Where Can We Fool the Model's Mind? Response-Guided Semantic Probing and Adaptive Routing for Transferable Attacks on Black-Box MLLMs27 Oct
- Combining Long-Term Memory and LLMs Comes at a High Cost: From Prompt Injection to Persistent Compromise20 Oct
- Same Simulation, Different Evidence: Where Retrieved Tool Output Costs LLM Agents Accuracy19 Oct
- AM-Bench: A Unified Taxonomy and Evaluation Suite for Agentic Misalignment16 Oct
- PhySimCode: A Benchmark and Evaluation Method for Physics Video to Code Generation11 Oct
- Eliciting Specifications: When Does Interactive Clarification Help?5 Oct
- The Pushback Paradox: A Two-Probe Diagnostic for Language Model Compliance5 Oct
- SpatialChain: A Benchmark for Auditing Spatial Reasoning Faithfulness in VLMs5 Oct
Same family
Claude Sonnet 4Alternatives
other vendors · similar capability- 29.7
- 33.6
- 32.5
- 30.4
Want Claude Sonnet 4.6 on OSMEXO?
Sign in to request it. Requests are reviewed; a request does not guarantee availability.









