Cheaper overall
GPT-6 Sol
80% lower blended rate
Higher benchmark scores
GPT-6 Astra
leads on 6 of 8 figures
Cheaper cached input
GPT-6 Sol
$0.20 vs $1.00, 80% less
Where each one wins
GPT-6 Sol
- Cheaper input tokens $2.00 vs $10.00, 80% less
- Cheaper output tokens $10.00 vs $50.00, 80% less
- Cheaper cached input $0.20 vs $1.00, 80% less
- Cheaper on all four workloads
- Higher long-context reasoning (AA-LCR) 83.7 vs 80.7
- Higher scientific coding (SciCode) 57.6 vs 56.5
GPT-6 Astra
- Higher overall intelligence (Intelligence Index) 52.7 vs 47.5
- Higher expert-exam performance (Humanity's Last Exam) 54.7 vs 47.9
- Higher physics research reasoning (CritPt) 31.7 vs 30.9
- Ahead on 3 more benchmark figures
- Offers Batch, Flex and Priority pricing not listed for GPT-6 Sol
What four workloads cost
One run, and the same run a thousand times.
GPT-6 Sol is cheaper on all four workloads, by much the same margin each time — about 5 times, from $0.0070 against $0.0350 on a chat turn to $1.65 against $8.23 on a long document.
| Workload | GPT-6 Sol | GPT-6 Astra | ×1,000 | Difference |
|---|---|---|---|---|
| Chat turn 1,000 in / 500 out | $0.0070 | $0.0350 | $7.00 / $35.00 | GPT-6 Sol is 80% cheaper |
| RAG answer 10,000 in / 800 out | $0.0280 | $0.1400 | $28.00 / $140.00 | GPT-6 Sol is 80% cheaper |
| Agent loop 32,000 in / 700 out × 20 calls | $0.3400 | $1.70 | $340.00 / $1,700.00 | GPT-6 Sol is 80% cheaper |
| Long document 400,000 in / 3,000 out | $1.65 | $8.23 | $1,645.00 / $8,225.00 | GPT-6 Sol is 80% cheaper |
Price per 1M tokens
Above 272,000 input tokens the rates change, and the new rate applies to the whole request, not just the tokens past that point: GPT-6 Sol's input goes from $2.00 to $4.00, and GPT-6 Astra's input goes from $10.00 to $20.00. It does not change which of the two is cheaper.
| Context band | Price | GPT-6 Sol | GPT-6 Astra | Difference |
|---|---|---|---|---|
| Prompts up to 272,000 tokens | Input | $2.00 | $10.00 | GPT-6 Sol is 80% 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 | $50.00 | GPT-6 Sol is 80% cheaper |
| Prompts over 272,000 tokens | Input | $4.00 | $20.00 | GPT-6 Sol is 80% cheaper |
| Prompts over 272,000 tokens | Cached input | $0.40 | $2.00 | GPT-6 Sol is 80% cheaper |
| Prompts over 272,000 tokens | Output | $15.00 | $75.00 | GPT-6 Sol is 80% cheaper |
Blended: GPT-6 Sol $4.00, GPT-6 Astra $20.00 per 1M — one rate at 3:1 input to output, for comparing two models at a glance.
GPT-6 Sol priced by Perplexity, GPT-6 Astra by OpenAI. Standard tier, pay-as-you-go.
Other pricing tiers
Only GPT-6 Astra lists Batch, Flex and Priority, at up to 50% off its own standard rate; GPT-6 Sol offers none of them.
Every tier either model sells is here, so this is the whole price sheet. A saving is measured against that model's own standard rate — how we read tiers.
| Tier | GPT-6 Sol | GPT-6 Astra | Cheaper |
|---|---|---|---|
| Batch only on GPT-6 Astra | Not offered | $5.00 in / $25.00 out 50% off Standard | — |
| Flex only on GPT-6 Astra | Not offered | $5.00 in / $25.00 out 50% off Standard | — |
| Priority only on GPT-6 Astra | Not offered | $20.00 in / $100.00 out | — |
Benchmarks
They share 8 figures: GPT-6 Sol leads on 2, GPT-6 Astra on 6. The widest gap is 8.8 points, on answer reliability (Omniscience Non-Hallucination).
Overall intelligence
Intelligence Index — Overall intelligence across reasoning, knowledge & math evals
GPT-6 Astra ahead by 5.2
Expert-exam performance
Humanity's Last Exam — Expert-level questions across many academic domains
GPT-6 Astra ahead by 6.8
Long-context reasoning
AA-LCR — Long-context reasoning across large inputs
GPT-6 Sol ahead by 3
Scientific coding
SciCode — Research-level scientific coding tasks
GPT-6 Sol ahead by 1.1
Physics research reasoning
CritPt — Unpublished physics research reasoning problems
GPT-6 Astra ahead by 0.8
Real-world task performance
GDPval — Economically valuable, real-world knowledge work
GPT-6 Astra ahead by 2.8
Factual accuracy
Omniscience Accuracy — Breadth of factual knowledge across domains
GPT-6 Astra ahead by 8.1
Answer reliability
Omniscience Non-Hallucination — How reliably the model avoids fabricated answers
GPT-6 Astra ahead by 8.8
Each figure is that model's best published run, with the effort level named beside it — where these scores come from.
Specifications
Only GPT-6 Astra lists web search.
| Specification | GPT-6 Sol | GPT-6 Astra |
|---|---|---|
| Context window | 1,050,000 tokens | 1,050,000 tokens |
| Max output | 128,000 tokens | 128,000 tokens |
| Takes in | Text, image, file | Text, image, file |
| Puts out | Text | Text |
| Knowledge cutoff | 20 Apr 2026 | 30 Apr 2026 |
| Released | 22 Sep 2026 | 3 Sep 2026 |
| Status | Active | Active |
| Sold by | Perplexity | OpenAI |
| Tool use | Yes | Yes |
| Structured outputs | Yes | Yes |
| Web search | Not listed | Yes |
| Prompt caching | Yes | Yes |
| Code execution | Yes | Yes |
| Computer use | Yes | Yes |
| Open weights | Not listed | Not listed |
| Reasoning | Yes | Yes |
FAQs
- Is GPT-6 Sol cheaper than GPT-6 Astra?
- Yes. A 1,000-token prompt with a 500-token reply costs $0.0070 on GPT-6 Sol and $0.0350 on GPT-6 Astra, and the same model is cheaper on every workload on this page.
- How much do GPT-6 Sol and GPT-6 Astra cost per 1M tokens?
- GPT-6 Sol costs $2.00 for input and $10.00 for output. GPT-6 Astra costs $10.00 and $50.00. Standard pay-as-you-go rates, as of 23 Sep 2026.
- Which is better, GPT-6 Sol or GPT-6 Astra?
- On published benchmarks, GPT-6 Astra. It leads on 6 of the 8 figures both models report, including overall intelligence (Intelligence Index), where it scores 52.7 against 47.5.
- Does GPT-6 Sol or GPT-6 Astra offer batch pricing?
- GPT-6 Astra only. Its batch input costs $5.00, 50% off its standard rate, and GPT-6 Sol lists no batch tier at all.