grok-4.6
The prior reasoning flagship from xAI, still highly capable for analytical and coding work with real-time X grounding. A strong choice when 4.7's newest gains aren't needed.
At a glance
Released
Aug 5, 2026
Provider
xAI
Context window
256,000 tokens
Max output
64,000 tokens
Modalities
Text + image → text
Knowledge cutoff
July 2026
API model ID
grok-4.6License
Proprietary (API only)
Pricing
Standard
Input$2 / 1M
Output$6 / 1M
Long-context requests are billed at $4 in / $12 out per 1M. A typical 1,500-token prompt with a 650-token answer costs ≈ $0.007 on the standard tier. Run your own prompt through the calculator →
What's new vs the previous model
Compared withgrok-4.5 (June 2026)
- Improved reasoning consistency over 4.5
- Edged out by 4.7 on the latest benchmarks
Best use cases
Analytical tasks
Handles complex reasoning and coding with dependable results.
Live-data queries
Grounds answers in real-time X content when needed.
Agent backends
Serves as a capable reasoning engine for tool-using agents.
Strengths & limitations
Strengths
- Capable reasoning
- Real-time grounding
- Stable behavior
Limitations
- Superseded by 4.7
- No audio or video input
How to test it
- No-code: open grok.com and pick grok-4.6 from the model switcher — no API key needed.
- Playground: sign in to the xAI Console and run prompts in the workbench, then copy the generated code.
- API: create a key in the xAI Console and call the API (example below).
curl https://api.x.ai/v1/chat/completions \
# OpenAI-compatible endpoint
-H "Authorization: Bearer $API_KEY" \
-H "Content-Type: application/json" \
-d '{
"model": "grok-4.6",
"messages": [{"role": "user", "content": "Explain tokenization in one paragraph."}]
}'