OpenAITextVisionAudio

gpt-5.6-sol

gpt-5.6-sol is OpenAI's model at $4.00 input and $20.00 output per million tokens, with a 1049K context window. On a typical 3:1 input-to-output mix that blends to roughly $8.00 per million tokens, placing it in the premium band of the 89 models tracked here — cheaper than 10% of them.

Open cost calculator View provider Open official platform

MODEL OVERVIEW

Context window1049K

Input price$4 / 1M

Output price$20 / 1M

Output speed85 tok/s

Quality score98

Data updatedUpdated Jul 2026

Price verified2026-08-22 · Official

By price, gpt-5.6-sol ranks 79th cheapest of 89 on input and 80th on output. Its quality score of 98 ranks 2nd of 89, and on value (quality score ÷ output price) it ranks 71st of 89. Its blended rate sits 160% above the catalog average.

In real billing terms: a support chatbot handling 50,000 turns a month (2,000 input / 500 output tokens each) costs about $900 per month on gpt-5.6-sol, while a RAG workload of 20,000 queries (8,000 input / 800 output tokens each) runs about $960. Because output is priced 5.0× input here, long-generation tasks scale cost fastest — cap output length before you optimise anything else.

A 1049K context window holds roughly 1,399 pages of A4 text (at ~750 tokens per page), ranking 8th of 89 here. That is enough to drop in an entire technical manual or several source files at once, which suits long-document summarization and codebase-wide analysis. Output runs at about 85 tok/s (53rd fastest of 89), acceptable interactively, though long responses will feel slow.

We track 15 OpenAI models, and gpt-5.6-sol's blended rate sits 71% above that provider's average. If your task tolerates a little less headroom, o3 (OpenAI) blends about 56% cheaper at a quality score of 95, and is worth benchmarking first. Going the other way, gpt-6-astra scores 99 on quality for about 150% more.

The closest model in OpenAI's line-up is gpt-5.4 ($2.50 input / $15.00 output, quality 92). Against it, gpt-5.6-sol scores 6 points higher on quality, runs 10 tok/s slower (about 11%), costs 30% more blended. When two models sit this close on price, the deciding factor is usually the quality ceiling — run both against your real prompts before picking.

Our read on it

After the 22 August cut, the highest-scoring model in the catalog is also the cheapest one scoring 96 or above.

$5/$30 became $4/$20, and that is a bigger move than it looks

On 22 August 2026 gpt-5.6-sol moved from $5/$30 to **$4/$20** against an official source — 20% off input, a third off output. Its quality score of 98 is the highest of the 68 models we track, and among everything scoring 96 or above — gpt-5.5 at $5/$30, claude-opus-5 at $5/$25, claude-fable-5 and claude-mythos-5 at $10/$50 — it is now the cheapest on both input and output.

That changes the argument you have to make internally. Previously, choosing Sol meant justifying the highest rate on the board — the work had to be compliance review, contract clause extraction, a whole-repository audit inside the 1,049K window, something where human verification costs far more than tokens. You no longer need that case. If you were going to run a flagship at all, Sol is the cheapest entry into that capability band, and no further argument is required.

The 85 tokens/second figure people skip past

Sol streams at 85 tokens/second. That is respectable for a flagship, but roughly half of luna's 150, and any interface where a user watches text appear will feel the difference. Sol belongs in asynchronous work — background batches, scheduled analysis, the planning step of an agent — where nobody is waiting on the first token.

Its 1,049K context across text, vision and audio pays off somewhere else entirely: you can often delete the chunking-and-retrieval layer from the pipeline instead of maintaining it. That saved engineering time tends to outweigh the token bill.

It now beats claude-opus-5 outright, which is not a reason to migrate tomorrow

Before the cut, claude-opus-5 at $5/$25 and quality 97 was 17% cheaper on output, and the two had genuinely different cases. At $4/$20 Sol is 20% cheaper on both sides, one point higher, matches the 1,049K context, and adds native audio that Opus 5 does not have. Every column now goes the same way.

Winning the table is not the same as being worth a migration. A composite quality score cannot see tool-use reliability, instruction adherence deep into a long conversation, or the fact that your prompts were tuned against one specific model. Moving between vendors means a full regression pass, and in the flagship band that usually costs more than a year of the price difference. Start new projects on Sol; run existing ones against shadow traffic before switching. And teams staying on OpenAI who cannot defend flagship pricing at all should look at gpt-5.6-terra — $2/$12, quality 95, 110 tokens/second: 40% cheaper and 30% faster than Sol, for three points.

Estimated monthly cost

Monthly API spend across four common workloads, next to the cheapest model we track.

Workload	Usage assumption	This model	Cheapest model
Support chatbot	2,000 input / 500 output tokens per turn, 50,000 turns per month	$900	$3.00
RAG document Q&A	8,000 input / 800 output tokens per query, 20,000 queries per month	$960	$3.84
Coding assistant	12,000 input / 2,000 output tokens per request, 10,000 requests per month	$880	$3.20
Batch summarization	4,000 input / 400 output tokens per item, 200,000 items per month	$4,800	$19

Derived directly from the rates on this page. Excludes prompt caching, batch discounts and free tiers.

How it compares

The alternatives worth benchmarking alongside it, from the same dataset.

Model	Provider	Blended /1M	Quality	Price delta
gpt-5.6-sol	OpenAI	$8.00	98	—
o3Cheaper option	OpenAI	$3.50	95	-56%
gpt-6-astraHigher quality	OpenAI	$20.00	99	+150%
gpt-5.4Same provider	OpenAI	$5.63	92	-30%
qwen3.5Best value overall	Alibaba	$0.11	91	-99%

Best for

ReasoningCodingMultimodal

Modalities

TextVisionAudio

Where it fits — and where it doesn't

Good fit

Coding: writing, refactoring and debugging across multi-file project context.
Reasoning: multi-step decomposition, planning and tasks that need rigorous inference.
Long documents: a 1049K window ingests a full contract, paper or large codebase in one pass.
Multimodal input: handles images, screenshots and scanned documents directly.

Poor fit

High-frequency batch jobs: at $8.00 blended per million tokens, bills scale quickly at volume.

Frequently asked questions

How much does the gpt-5.6-sol API cost?

gpt-5.6-sol costs $4.00 per million input tokens and $20.00 per million output tokens. On a 3:1 input-to-output mix that blends to about $8.00 per million tokens.

What does gpt-5.6-sol cost per month?

It depends on volume. Using the four worked scenarios on this page: a support chatbot runs about $900/month, RAG document Q&A about $960/month, a coding assistant about $880/month, and batch summarization about $4,800/month. Plug your own token counts into the cost calculator for a tailored figure.

How large is the gpt-5.6-sol context window?

1049K tokens — roughly 1,399 pages of A4 text at ~750 tokens per page, ranking 8th of the 89 models tracked here.

gpt-5.6-sol or o3 — which is better value?

o3 blends to about $3.50 per million tokens versus $8.00 for gpt-5.6-sol — roughly 56% cheaper — at a quality score of 95 against 98. If that quality gap does not show up on your task, take the cheaper one; if it does, stay with gpt-5.6-sol. Benchmark both on your own data before deciding.

What is gpt-5.6-sol best suited for?

Tagged use cases: reasoning, coding, multimodal. Coding: writing, refactoring and debugging across multi-file project context. Less suitable: High-frequency batch jobs: at $8.00 blended per million tokens, bills scale quickly at volume.

How current is this gpt-5.6-sol pricing?

Pricing here was last verified on 2026-08-22 against official documentation. llmprice.app runs a daily collection job, but providers can change rates between runs — confirm against OpenAI's official pricing page before you commit.