Grok 4.6
xai/grok-4.6- xAI
- Family · Grok 4
- Released 12 Aug 2026
- Text · Image · File → Text
- Reasoning · adjustable
- Tools · JSON · Vision
Not on OSMEXO yet
Spec & list price
Context488K
Max output450K
Input / 1M$2.00
Output / 1M$6.003× input
Scores · standing
Intelligence44.3#22/489
LiveBench67.8#15/32
Coding76.8#5/196
Usage
Tokens / day4.1B▼ 52% 7d
Usage rank#125
Requests / day129.1K
Tool calls / req0.57
Per request · 30 days
Avg input34.4K2.3× median
Avg output1.5K2.2× median
Cache hit80%1.1× median
Reasoning share82%1.5× median
Tool-call failures0.3%0.1× median
Consumption
Peak 36.8B 30 SepLow 3.2B 1 NovRank today #125
News & trends
40 mentions this week ▲ 186%Grok 4 family · videos, posts, articles · 30 days
Trending · 14 days · Grok 4 family
- Rubik's Cube Solves Itself - Gemini 4 Argon vs Grok 4.7 vs GPT-6 Astra vs Claude Opus 5.51.4K views
- Grok 4.3 in bad. Where do you rank GLM 5.3 Flash?14 upvotes
- DeepSeek 4.1 Flash ended up at the bottom of Do the Work. Where would you rank Grok 4.3?11 upvotes
- Kimi K3.1 Pro vs Claude Opus 5.5 vs Grok 4.7 Ai Comparison 2026926 views
- @OpenRouter on X18.8K views
Benchmarks
bar = share of the top score · Benchmark data: Artificial AnalysisSpeed · median across 4 providers, latest day
Output speed#198/33458 tok/s1.1× slower than median
Fastest provider90 tok/sbest of 4 providers
First token#2561.89 s1.8× longer than median
#8 cheapest of 13 with a similar score
Value
Intelligence Index
Blended price per 1M tokens (log)
Family progress
Intelligence Index
- Family releases46.4
LiveBench suites
Browser49.4
Coding89.1
Data Analysis75.3
Design79.8
Long Agentic28.0
Reasoning91.2
Research84.5
▲ Strongest: Reasoning, Coding. ▼ Weakest: Long Agentic (28).
Technical details
- Reasoning
- on · high
- Tokenizer
- Grok
- Moderation
- none
Reasoning effort
- xhigh
- high · default
- medium
- low
Request parameters 15
- tools
- tool_choice
- response_format
- structured_outputs
- reasoning
- reasoning_effort
- include_reasoning
- temperature
- top_p
- top_k
+5 more
- max_tokens
- stop
- seed
- logprobs
- top_logprobs
Interesting facts
- 4.7× the output room of similar models (450.0K against a 96.8K median).
- 2nd of 5 Grok 4 models by tokens last month.
When to use
Best for
- Hard reasoning and graduate-level science
- Coding and terminal agent tasks
- Browser and research agents
- Customer-service style tool workflows
Think twice for
- Self-hosting (closed weights)
- Low-cost high-volume chat
- Audio input
Research papers
- Being judged through the record: A Rosenhan-inspired test of psychiatric label sensitivity in large language models26 Oct
- MolHop: Scaffold-Hopping Molecular Design with Agentic Reinforcement Learning14 Oct
- Autonomous Active Directory Exploitation via Multi-Model Harness Orchestration: A Benchmark Study with NeuroSploit on GOAD3 Oct
- DrivingBench: Can Vision-Language Models Drive a Toyota Corolla?30 Sep
- Playing log(N)-Questions over Wikipedia Abstracts: How Per-Round Errors Compound Under Information Asymmetry16 Sep
- Fröberg's Conjecture for Quintics and Septics in Four Variables25 Aug
Social
Most watched and discussed · 90 days
- GPT 5.6 and Grok 4.6 models added to Brave's Leo AI assistantPiunikaWeb - your daily source of web browser news · 65 views · 24 Oct
Same family
Grok 4- 25.2
- 24.9
- 46.4
- 38.8
- 22.5
Alternatives
other vendors · similar capability- S43.7
- 40.9
- 47.0
- 47.6
Want Grok 4.6 on OSMEXO?
Sign in to request it. Requests are reviewed; a request does not guarantee availability.