💰 Promo pricing — $5→$4 / $30→$20 (at least through Nov 21, 2026)

GPT-5.6 Sol

OpenAI's current flagship for complex professional work and long-horizon agents. At the August promo price, now cheaper than Claude Opus 5.

$4.00
Input / 1M tokens (promo)
$20.00
Output / 1M tokens (promo)
1.05M
Context window
128K
Max output tokens
Last updated: | Sources: OpenAI Model Docs · OpenAI Pricing · OpenAI announcement (8/21) | All Pricing · Calculator · 日本語版

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).

ModeInputCached InputCache WritesOutputNotes
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$10.0050% off Standard. Up to 24h
Flex$2.00$0.20$2.50$10.00Lower priority. Same as Batch
Fast mode$8.00$0.80$10.00$40.00Priority. Short context
Price history:
$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 The post-promo rate is unpublished — do not budget past Nov 21 on these figures.
☁️ Cloudflare AI Gateway: limited-time 50% off (through Sep 18, 2026) — per Cloudflare's changelog (2026-08-19), routing to 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

GA Release
July 9, 2026
Knowledge Cutoff
Feb 16, 2026
Context Window
1,050,000
Max Output
128,000
Input Modalities
Text + Image
Output Modalities
Text
Reasoning Effort
none–max (6 levels)
Tool Use
Function Calling ✓
Fine-tuning
Not supported
Model ID
gpt-5.6-sol

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.effort levels 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.

ModelProviderInputOutputvs SolNotes
GPT-5.6 SolOpenAI$4.00$20.00Promo (through 11/21)
Claude Fable 5Anthropic$10.00$50.00Sol cheaperSol: 60% cheaper input & output
Claude Opus 5Anthropic$5.00$25.00Sol cheaperSol: 20% cheaper input & output
GPT-5.5OpenAI$5.00$30.00Sol cheaperPrevious flagship
GPT-5.6 LunaOpenAI$0.20$1.20Luna cheaperLightweight tier. Detail page
Grok 4.6xAI$2.00$6.00Grok cheaperAA Index 61, reportedly on par with Sol
DeepSeek V4 ProDeepSeek$0.66–$1.32$1.98–$3.96DeepSeek cheaperPeak/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

ModelInputOutputWhen to Use
GPT-5.6 Sol ⬅$4.00$20.00Hardest tasks, long-horizon agents, high-failure-cost work
GPT-5.6 Terra$2.00$12.00Standard coding, analysis, production agents
GPT-5.6 Luna$0.20$1.20High-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.