Pricing (promotional — verified 2026-08-29)
Verified against OpenAI's official pricing page (developers.openai.com). USD per 1M tokens. Promotional pricing is valid at least through November 21, 2026 (stated by OpenAI).
| Mode | Input | Cached Input | Cache Writes | Output | Notes |
|---|---|---|---|---|---|
| Standard (Short context) | $4.00 | $0.40 | $5.00 | $20.00 | ≤272K input tokens. Default rate |
| Standard (Long context) | $8.00 | $0.80 | $10.00 | $30.00 | >272K input tokens. 2× input, 1.5× output |
| Batch | $2.00 | $0.20 | $2.50 | $10.00 | 50% off Standard. Up to 24h |
| Flex | $2.00 | $0.20 | $2.50 | $10.00 | Lower priority. Same as Batch |
| Fast mode | $8.00 | $0.80 | $10.00 | $40.00 | Priority. Short context |
$5.00 / $30.00 (standard at GA launch, July 9, 2026)
→ $4.00 / $20.00 (promotional, announced Aug 21, 2026 — 20% off input, 33% off output)
* The promo rate is available "at least through November 21, 2026." The post-promo rate is unpublished — do not budget past Nov 21 on these figures.
openai/gpt-5.6-sol applies the discount automatically. Unified Billing users only (Bring Your Own Keys excluded). Cloudflare's published promo rates: input $2.50 / output $15 / cache read $0.25 per 1M (vs Cloudflare's listed $5/$30/$0.50 standard). Cloudflare's announcement does not establish that this stacks with OpenAI's direct $4/$20 promotion — confirm the actual billed amount in the Cloudflare dashboard.
Primary sources: developers.openai.com/api/docs/pricing · OpenAI Developer Community announcement (2026-08-21) · Cloudflare Changelog (2026-08-19) (all verified 2026-08-29)
Model Specifications
Source: OpenAI GPT-5.6 Sol Model Docs (verified 2026-08-29). The gpt-5.6 alias routes to Sol. No audio/video (image input only).
Performance & Positioning
- Official positioning: "Flagship model for complex professional work." Highest reasoning tier with
reasoning.effortlevels none / low / medium (default) / high / xhigh / max. - Full Responses API tooling: web search, file search, image generation, code interpreter, hosted shell, apply patch, skills, computer use, MCP, tool search.
- vs the previous flagship (input): GPT-5.5 $5.00 → Sol $4.00, same-tier intelligence at a lower price.
- 1.05M context + 128K output for very large repositories and long-horizon agent tasks.
Benchmark figures change rapidly; this page lists only specs and pricing verified against primary sources. For quantitative benchmarks, refer to evaluators such as Artificial Analysis.
Flagship Price Comparison
USD per 1M tokens (standard, short context). Sol's promo price makes it a strong choice in the top tier.
| Model | Provider | Input | Output | vs Sol | Notes |
|---|---|---|---|---|---|
| GPT-5.6 Sol | OpenAI | $4.00 | $20.00 | — | Promo (through 11/21) |
| Claude Fable 5 | Anthropic | $10.00 | $50.00 | Sol cheaper | Sol: 60% cheaper input & output |
| Claude Opus 5 | Anthropic | $5.00 | $25.00 | Sol cheaper | Sol: 20% cheaper input & output |
| GPT-5.5 | OpenAI | $5.00 | $30.00 | Sol cheaper | Previous flagship |
| GPT-5.6 Luna | OpenAI | $0.20 | $1.20 | Luna cheaper | Lightweight tier. Detail page |
| Grok 4.6 | xAI | $2.00 | $6.00 | Grok cheaper | AA Index 61, reportedly on par with Sol |
| DeepSeek V4 Pro | DeepSeek | $0.66–$1.32 | $1.98–$3.96 | DeepSeek cheaper | Peak/off-peak. See China AI overview |
Sources: official pricing pages (verified 2026-08-29). See the full pricing comparison.
Ideal Use Cases
For high-stakes workloads that demand top-tier reasoning.
Large-scale codebase work
1.05M context + 128K output for cross-repository understanding and refactoring.
Long-horizon autonomous agents
Computer use, MCP and tool calling for multi-step automation.
Research & hard problems
Max reasoning effort for expert-level, difficult tasks.
Professional work (finance, legal, medical)
Where errors are expensive, top-tier accuracy pays for itself (support/drafting use — final calls need expert review).
Batch & Flex processing
Another 50% off via Batch/Flex. For overnight jobs.
RAG & search-augmented gen
High-accuracy answers with web search + file search tools.
GPT-5.6 Family Selection Guide
| Model | Input | Output | When to Use |
|---|---|---|---|
| GPT-5.6 Sol ⬅ | $4.00 | $20.00 | Hardest tasks, long-horizon agents, high-failure-cost work |
| GPT-5.6 Terra | $2.00 | $12.00 | Standard coding, analysis, production agents |
| GPT-5.6 Luna | $0.20 | $1.20 | High-throughput, cost-sensitive, routine work, first-pass review |
💡 Strategy: start with Luna, escalate to Terra→Sol when quality is insufficient. Reserve Sol for the final high-accuracy pass and hard tasks.
Estimate Your Actual Costs
Compare token costs for GPT-5.6 Sol (promo) and all major models. Just enter your token counts.
🧮 Open Token Calculator →📊 Full Pricing Comparison · ⚡ Compare with GPT-5.6 Luna · 🗓️ Release Timeline · 🇨🇳 China AI Overview · 🔗 OpenAI Official Pricing
⚠️ Disclaimer
- Pricing verified against OpenAI's official page (developers.openai.com) on 2026-08-29. The promotional rate ($4/$20) is stated to be available "at least through November 21, 2026"; the post-promo rate is unpublished. Always check official pages before contracting.
- Cloudflare AI Gateway's 50% off runs through 2026-09-18 and is for Unified Billing users only (Bring Your Own Keys excluded).
- This site is informational only. We are not affiliated with or endorsed by any provider.