Cheaper overall
Grok-4.7
25% lower blended rate
Cheaper on long prompts
GPT-6 Sol
above 199,999 input tokens
Higher benchmark scores
GPT-6 Sol
leads on 6 of 8 figures
Bigger context window
GPT-6 Sol
1,050,000 vs 500,000 tokens
Cheaper cached input
GPT-6 Sol
$0.20 vs $0.50, 60% less
Where each one wins
GPT-6 Sol
- Cheaper cached input $0.20 vs $0.50, 60% less
- Cheaper on long prompts above 199,999 input tokens
- Larger context window 1,050,000 tokens
- Higher overall intelligence (Intelligence Index) 47.5 vs 46.4
- Higher expert-exam performance (Humanity's Last Exam) 47.9 vs 43.1
- Higher long-context reasoning (AA-LCR) 83.7 vs 76.7
- Ahead on 3 more benchmark figures
Grok-4.7
- Cheaper output tokens $6.00 vs $10.00, 40% less
- Higher real-world task performance (GDPval) 59.8 vs 49.3
- Higher answer reliability (Omniscience Non-Hallucination) 70.7 vs 39.9
What four workloads cost
One run, and the same run a thousand times.
Caching is what decides it: over twenty calls GPT-6 Sol costs $0.3400 against $0.4640, a difference of $124.00 across a thousand loops.
| Workload | GPT-6 Sol | Grok-4.7 | ×1,000 | Difference |
|---|---|---|---|---|
| Chat turn 1,000 in / 500 out | $0.0070 | $0.0050 | $7.00 / $5.00 | Grok-4.7 is 29% cheaper |
| RAG answer 10,000 in / 800 out | $0.0280 | $0.0248 | $28.00 / $24.80 | Grok-4.7 is 11% cheaper |
| Agent loop 32,000 in / 700 out × 20 calls | $0.3400 | $0.4640 | $340.00 / $464.00 | GPT-6 Sol is 27% cheaper |
| Long document 400,000 in / 3,000 out | $1.65 | $1.64 | $1,645.00 / $1,636.00 | Grok-4.7 is 0.5% cheaper |
Price per 1M tokens
Above 199,999 input tokens the rates change, and the new rate applies to the whole request, not just the tokens past that point: Grok-4.7's input goes from $2.00 to $4.00. That is where the cheaper model changes.
| Context band | Price | GPT-6 Sol | Grok-4.7 | Difference |
|---|---|---|---|---|
| Prompts up to 199,999 tokens | Input | $2.00 | $2.00 | Same |
| Prompts up to 199,999 tokens | Cached input | $0.20 | $0.50 | GPT-6 Sol is 60% cheaper |
| Prompts up to 199,999 tokens | Output | $10.00 | $6.00 | Grok-4.7 is 40% cheaper |
| Prompts up to 200,000 tokens | Input | $2.00 | $4.00 | GPT-6 Sol is 50% cheaper |
| Prompts up to 200,000 tokens | Cached input | $0.20 | $1.00 | GPT-6 Sol is 80% cheaper |
| Prompts up to 200,000 tokens | Output | $10.00 | $12.00 | GPT-6 Sol is 17% cheaper |
| Prompts up to 272,000 tokens | Input | $2.00 | $4.00 | GPT-6 Sol is 50% cheaper |
| Prompts up to 272,000 tokens | Cached input | $0.20 | $1.00 | GPT-6 Sol is 80% cheaper |
| Prompts up to 272,000 tokens | Output | $10.00 | $12.00 | GPT-6 Sol is 17% cheaper |
| Prompts over 272,000 tokens | Input | $4.00 | $4.00 | Same |
| Prompts over 272,000 tokens | Cached input | $0.40 | $1.00 | GPT-6 Sol is 60% cheaper |
| Prompts over 272,000 tokens | Output | $15.00 | $12.00 | Grok-4.7 is 20% cheaper |
Blended: GPT-6 Sol $4.00, Grok-4.7 $3.00 per 1M — one rate at 3:1 input to output, for comparing two models at a glance.
Both priced by Perplexity. Standard tier, pay-as-you-go.
Benchmarks
They share 8 figures: GPT-6 Sol leads on 6, Grok-4.7 on 2. The widest gap is 30.8 points, on answer reliability (Omniscience Non-Hallucination).
Overall intelligence
Intelligence Index — Overall intelligence across reasoning, knowledge & math evals
GPT-6 Sol ahead by 1.1
Expert-exam performance
Humanity's Last Exam — Expert-level questions across many academic domains
GPT-6 Sol ahead by 4.8
Long-context reasoning
AA-LCR — Long-context reasoning across large inputs
GPT-6 Sol ahead by 7
Scientific coding
SciCode — Research-level scientific coding tasks
GPT-6 Sol ahead by 0.2
Physics research reasoning
CritPt — Unpublished physics research reasoning problems
GPT-6 Sol ahead by 13.2
Real-world task performance
GDPval — Economically valuable, real-world knowledge work
Grok-4.7 ahead by 10.5
Factual accuracy
Omniscience Accuracy — Breadth of factual knowledge across domains
GPT-6 Sol ahead by 7.1
Answer reliability
Omniscience Non-Hallucination — How reliably the model avoids fabricated answers
Grok-4.7 ahead by 30.8
Each figure is that model's best published run, with the effort level named beside it — where these scores come from.
Specifications
GPT-6 Sol holds 550,000 more input tokens in one request. Only GPT-6 Sol lists computer use.
| Specification | GPT-6 Sol | Grok-4.7 |
|---|---|---|
| Context window | 1,050,000 tokens | 500,000 tokens |
| Max output | 128,000 tokens | — |
| Takes in | Text, image, file | Text, image, file |
| Puts out | Text | Text |
| Knowledge cutoff | 20 Apr 2026 | 31 May 2026 |
| Released | 22 Sep 2026 | 21 Sep 2026 |
| Status | Active | Active |
| Sold by | Perplexity | Perplexity |
| Tool use | Yes | Yes |
| Structured outputs | Yes | Yes |
| Web search | Not listed | Not listed |
| Prompt caching | Yes | Yes |
| Code execution | Yes | Yes |
| Computer use | Yes | Not listed |
| Open weights | Not listed | Not listed |
| Reasoning | Yes | Yes |
FAQs
- Is GPT-6 Sol cheaper than Grok-4.7?
- No — the other way round. A 1,000-token prompt with a 500-token reply costs $0.0070 on GPT-6 Sol and $0.0050 on Grok-4.7.
- How much do GPT-6 Sol and Grok-4.7 cost per 1M tokens?
- GPT-6 Sol costs $2.00 for input and $10.00 for output. Grok-4.7 costs $2.00 and $6.00. Standard pay-as-you-go rates, as of 23 Sep 2026.
- Which is better, GPT-6 Sol or Grok-4.7?
- On published benchmarks, GPT-6 Sol. It leads on 6 of the 8 figures both models report, including overall intelligence (Intelligence Index), where it scores 47.5 against 46.4.
- Is GPT-6 Sol or Grok-4.7 cheaper for long prompts?
- GPT-6 Sol, above 199,999 input tokens. A rate that changes at a context length applies to the whole request, not only the tokens past it, so the switch is sharp rather than gradual.
- Does GPT-6 Sol or Grok-4.7 have a bigger context window?
- GPT-6 Sol, at 1,050,000 tokens against 500,000. That is 550,000 more input tokens in a single request.