Models

Claude Sonnet 4.6 Thinking vs Grok 4.5

Output tokens cost $15.00 per million on Claude Sonnet 4.6 Thinking and $6.00 per million on Grok 4.5. Input tokens cost $3.00 per million on Claude Sonnet 4.6 Thinking and $2.00 per million on Grok 4.5. Cached input tokens are billed at $0.30 per million on Claude Sonnet 4.6 Thinking and $0.50 per million on Grok 4.5. Context lengths are 200,000 tokens on Claude Sonnet 4.6 Thinking and 500,000 tokens on Grok 4.5. Time to first token (TTFT) measured on AIHubMix is 1.2s on Claude Sonnet 4.6 Thinking and 2.2s on Grok 4.5. Measured output throughput is 67.8 tok/s on Claude Sonnet 4.6 Thinking and 27.1 tok/s on Grok 4.5. On the LMArena coding leaderboard Claude Sonnet 4.6 Thinking scores 1528 and Grok 4.5 scores 1523.

AnthropicClaude Sonnet 4.6 ThinkingGrokGrok 4.5
Anthropic logo
Claude Sonnet 4.6 Thinking
Anthropic · image, text → text

Claude sonnet 4.6 does not enable reasoning mode by default. To access its deep reasoning capabilities, users would typically need to call the native Claude API. To make this capability available through an OpenAI-compatible interface, we provide the claude-sonnet-4-6-think model, which has reasoning mode pre-enabled and a default 32k-token context window, allowing it to be called directly via the OpenAI unified API. Claude sonnet 4.5 Think is a reasoning-focused variant of Claude sonnet 4.6 designed for advanced tasks that require rigorous reasoning, complex decision-making, and long-chain analysis. Aside from its enhanced reasoning mechanism, all other capabilities remain consistent with the standard Claude sonnet 4.6 model, making it well suited for complex engineering problem decomposition, multi-stage planning, and logic-intensive analysis.

Input$3.00 /M
Output$15.00 /M
Grok logo
Grok 4.5
Grok · text, image → text

Grok 4.5 was trained on datasets spanning knowledge in coding, science, engineering, and math. With both intelligent and efficient reasoning, Grok 4.5 excels at real engineering tasks and exceeds comparable leading models at these tasks.

Input$2.00 /M
Output$6.00 /M

Pricing & Specifications

Prices are per million tokens. Time to First Token and throughput are rolling averages measured on AIHubMix.

Claude Sonnet 4.6 Thinking
Grok 4.5
Input /M
$3.00
$2.00
Output /M
$15.00
$6.00
Cache read /M
$0.30
$0.50
Context length
200,000
500,000
Max output
32,000
500,000
Time to First Token
1.2 s
2.2 s
Throughput
67.8 tok/s
27.1 tok/s
Modalities
imagetext
textimage
Supported Parameters
thinkingtoolsfunction callingstructured outputs
thinkingtoolsfunction callingstructured outputslong context
API Formats
Released
-
July 2026

Promotional prices show the discounted rate; see each model page for promotion windows.

Activity Past 30 Days

Daily traffic served through AIHubMix — how demand for each model is trending.

claude-sonnet-4-6-thinkgrok-4.5

Tokens / day

-

Requests / day

-

Performance Past 3 Days

Measured on real AIHubMix traffic, hourly buckets. Gaps mean no traffic in that hour.

claude-sonnet-4-6-thinkgrok-4.5

Throughput (tok/s)

-

TTFT (s)

-

Uptime (%)

-

LMArena Benchmarks

LMArena ratings by capability (Bradley-Terry, commonly called Elo). Higher is better.

Text
claude-sonnet-4-6-thinkgrok-4.5
1420146015001540
Overall
14701472
Coding
15231528
Math
14631478
Hard prompts
14951503
Instruction following
14661476
Multi-turn
14771481
Creative writing
14491450
Longer query
14851493
Chinese
15101511
English
14751484
Vision
claude-sonnet-4-6-thinkgrok-4.5
1240128013201360
Overall
12751282
OCR
12901295
Diagram
13061321
Homework
13021328
WebDev Arena
claude-sonnet-4-6-thinkgrok-4.5
1480152015601600
Overall
15221556
React
15121559
HTML
15291577
Gaming
15421586
Simulations
15221598
Data analytics
15191522

Source: LMArena (arena.ai) leaderboard, imported by AIHubMix. Models without published ratings are omitted per chart.

Cost calculator

Estimate your monthly bill for the same workload on each model.

Grok 4.5
$210 /mo
Claude Sonnet 4.6 Thinking
$405 /mo

Monthly = daily × 30. Discounted rates applied where a promotion is active.

FAQ

Which is cheaper: Claude Sonnet 4.6 Thinking, Grok 4.5?

Grok 4.5: $6.00/M output tokens; Claude Sonnet 4.6 Thinking: $15.00/M. Use the cost calculator above to estimate your own workload.

How do their coding arena scores compare?

Claude Sonnet 4.6 Thinking: 1528; Grok 4.5: 1523 (LMArena coding leaderboard).

Which responds faster?

Claude Sonnet 4.6 Thinking: 1.2s time to first token measured on AIHubMix; see the live performance charts above for how each model behaves across the day.

How large is each context window?

Claude Sonnet 4.6 Thinking accepts 200,000 and Grok 4.5 accepts 500,000 input tokens. Maximum output per request is 32,000 tokens on Claude Sonnet 4.6 Thinking and 500,000 tokens on Grok 4.5.

Which one generates tokens faster?

Claude Sonnet 4.6 Thinking at 67.8 tok/s and Grok 4.5 at 27.1 tok/s, measured as output throughput on AIHubMix — a separate metric from time to first token.

What inputs and capabilities does each model support?

Claude Sonnet 4.6 Thinking accepts image and text input and supports thinking, tool calling, function calling and structured outputs; Grok 4.5 accepts text and image input and supports thinking, tool calling, function calling, structured outputs and long context.

Can I call Claude Sonnet 4.6 Thinking and Grok 4.5 with the same API key?

Yes. AIHubMix serves every model on this page behind one OpenAI-compatible endpoint, so switching between them is a one-line change to the model field — no second account, key or SDK.

Popular comparisons

Related model match-ups readers also look at.