Opus 5.5 and GPT-6 Sol Cut Prices the Same Day

On 22 September 2026 Anthropic shipped Claude Opus 5.5 and cut its price 20%. About an hour later OpenAI shipped GPT-6 Sol and called it 50% cheaper. Both headlines are true, but they use different baselines, so neither helps you pick what to run.

Put both on the same job instead: 1M input tokens and 100K output with no caching. GPT-6 Sol costs $3.00 per request and Claude Opus 5.5 costs $6.00. Sol is half the price.

Cross 272,000 input tokens and the gap nearly closes. Above that line Sol bills at $4 per million input, exactly Opus 5.5's rate, and its cached input rate becomes worse than Anthropic's.

Table of contents

• What shipped, and when
• Where Anthropic's 40% comes from
• OpenAI's 50% and its baseline
• Cost on identical work
• The 272K billing threshold
• Benchmarks
• Moving from Opus 5 to Opus 5.5
• Videos API shutdown
• FAQ
• Sources

What shipped, and when

Three labs shipped frontier text models inside about 36 hours:

Two flagship price cuts landed on the same day, and the percentages attached to them can't be compared. Both labs were fairly upfront about why.

Where Anthropic's 40% comes from

Anthropic's headline is a 40% cost drop versus Opus 5. Its own wording gives the arithmetic: Opus 5.5 "costs less per token than Opus 5 and uses fewer tokens per task, which nets out to a 40% drop in costs."

That is two effects stacked. The per-token half:

The cache read cut is the biggest single change, and most coverage skips it. On a job with 1M input at a 90% cache hit rate plus 100K output, Opus 5 costs $3.45 and Opus 5.5 costs $2.58. That is a 25.2% saving on price alone.

The other 15 points come from token efficiency. Anthropic measured it on typical agentic tasks at default settings, and the figure is honest for that. It still depends on your workload, so it can't go into a budget spreadsheet the way $4/MTok can.

Opus 5.5 also has a Fast mode at $8/$40 for up to 2.5x speed, available in Claude Code and on the Claude Platform. That is double the standard rate. Anthropic says standard output already generates more than 30% faster than Opus 5.

OpenAI's 50% and its baseline

OpenAI's 50% is simple once you know the baseline. GPT-5.6 Sol is $4 input and $20 output, and GPT-6 Sol is $2 and $10. Exactly half.

But GPT-5.6 Sol's price is itself promotional, running through 21 November 2026 after a 20%-plus drop in August. The 50% is measured against a temporary rate, not a list price, so the absolute number tells you more than the percentage.

The absolute numbers for both new OpenAI models:

Luna is the one to look at if you run volume. Its $0.10 input rate is a fortieth of Opus 5.5's. OpenAI positions it as its most efficient model for focused, high-volume tasks, a narrower claim than Sol's "complex coding and agentic workflows."

Cost on identical work

Percentages off a moving baseline are noise. Cost per request on the same job is not. Here is 1M input plus 100K output, no caching:

With caching on at a 90% hit rate the ordering holds but the spread tightens. Opus 5.5 drops to $2.58, Opus 5 to $3.45, GPT-6 Sol to $1.38 and Fable 5.1 to $6.22.

Your token mix won't match mine. Output-heavy work punishes Opus 5.5 harder, since its output rate is double Sol's, and a long cached system prompt favours Anthropic. Run your own numbers through our API cost calculator, which carries all of these models at current rates.

The 272K billing threshold

Both GPT-6 models change how they bill above 272,000 input tokens. Past that line the entire request, not just the overage, bills at 2x input and cache rates and 1.5x output.

A 1M-input job goes from $3.00 to $5.50, an 83% increase, for a request one token over the threshold.

Effective rates on either side of the line, next to Opus 5.5:

Above the threshold Sol's input rate matches Opus 5.5's, and its cached input rate is twice Anthropic's. On that 1M-token job Sol's advantage shrinks from 50% cheaper to 8.3% cheaper, or 12.4% at a 90% cache hit rate. The surcharge applies to GPT-6 Luna too.

If your workload is long-context by nature (whole repositories, long transcripts, big document sets), the headline price isn't your price. Budget from the surcharged rate.

xAI does something similar at a lower threshold. Grok 4.7 costs $2 input, $0.50 cached and $6 output on a 500K context, same as Grok 4.6, and those rates double above 200K prompt tokens to $4 input, $1 cached and $12 output.

Benchmarks

Every number below is self-reported by Anthropic or run by a named partner. No independent head-to-head of Opus 5.5 against GPT-6 Sol exists yet.

GPT-6 Astra wins Terminal-Bench-Science, 64.6% against 58.7%. A post that lists only the wins is marketing.

The effort levels don't match either. Opus 5.5's Terminal-Bench figure is at xhigh effort and Astra's is at high effort, each model's own best reported score. Treating them as an equal compute budget overstates the gap.

On GDPval-AA v2.1, an Artificial Analysis evaluation across 44 occupations, Opus 5.5 scores 1846 Elo against Fable 5.1's 1735 and Opus 5's 1708.

Anthropic's own caveat is the most useful line in its announcement, so here it is verbatim: "at these levels of capability we've found that benchmark margins have become a less reliable guide to real-world differences. In our own use, the gap between Opus 5.5 and Claude Fable 5.1 is narrower than these scores suggest."

It also notes that safeguard interventions routed some evaluation tasks to older models, cybersecurity to Opus 4.8 and biology and frontier LLM development to Opus 5. Anthropic says this likely lowered Opus 5.5's own reported scores.

Moving from Opus 5 to Opus 5.5

Opus 5.5 is not a drop-in replacement. Anthropic's migration documentation lists four breaking changes:

1. Thinking cannot be disabled. If your code sets thinking off, that path is gone.
2. Forced tool use returns an error instead of complying.
3. Thinking blocks are tied to both the model and the conversation, so you can't move them across either.
4. On the Claude API and Google Cloud, the older computer_20251124 computer use tool is not accepted.

The first three apply to Claude Fable 5.1 as well, so a team already on Fable 5.1 has less work to do.

A fifth change fails no request, which makes it the dangerous one. Text that used to come back between tool calls now arrives inside thinking blocks whose text is empty at the default display setting. An app that streams that text as progress updates goes silent between tool calls. Nothing errors and nothing logs. The UI just stops talking until you set a display value that returns the text. If you ship an agent with a live progress feed, test that path before you change the model ID.

Once you've picked a model, the API cost calculator shows what the switch does to a month of your traffic, including the cached and long-context cases above. Four documented breaking changes and one silent behaviour change can cost more engineering hours than the 20% saves in the first month, so budget for the migration as well as the price.

Videos API shutdown

Separately, the Videos API shuts down on 24 September 2026, today. That covers sora-2, sora-2-pro and three dated snapshots.

Developers were notified on 24 March 2026, and the Sora web and app experiences ended on 26 April. The deprecations page lists a literal dash in the recommended replacement column, so there is no OpenAI video model to move to.

Anyone still calling it has to switch providers: Veo, Runway, Kling, Hailuo or Wan. All of them are in the calculator with per-second rates, which is the comparison you need because video pricing works nothing like token pricing.

FAQ

Is GPT-6 Sol cheaper than Claude Opus 5.5?

Yes, below 272K input tokens, by about half. On 1M input plus 100K output with no caching Sol costs $3.00 and Opus 5.5 costs $6.00. Above 272K input tokens Sol's surcharge takes the same job to $5.50, leaving it only 8.3% cheaper.

Is Claude Opus 5.5 really 40% cheaper than Opus 5?

Only partly. Anthropic's 40% combines a 20% per-token price cut with Opus 5.5 using fewer tokens per task. On per-token prices alone the saving is 20%, or 25.2% on a 1M-input job at a 90% cache hit rate because cache reads fell 60%.

Can I swap Opus 5 for Opus 5.5 without code changes?

No. Thinking cannot be disabled, forced tool use errors, thinking blocks are bound to the model and conversation, and the older computer_20251124 tool is rejected on the Claude API and Google Cloud. Text between tool calls also moves into thinking blocks that are empty at the default display setting.

What replaces the Sora 2 API?

Nothing from OpenAI. The deprecations page leaves the recommended replacement column empty. Veo, Runway, Kling, Hailuo and Wan are the working alternatives.

Sources

• Claude Opus 5.5 announcement: price cut, Fast mode, benchmark table and the caveat on benchmark margins
• Claude Opus 5.5 model documentation: rates, context window, cache pricing and the migration breaking changes
• GPT-6 Sol model reference: $2/$10 rates, 1.05M context and the 272K surcharge
• GPT-6 Luna model reference: $0.10/$0.50 rates and the same surcharge threshold
• xAI pricing: Grok 4.7 and Grok Build 0.1 rates and the 200K threshold
• OpenAI deprecations: Videos API, sora-2 and sora-2-pro shutdown on 2026-09-24 with no named replacement
• Cost-per-request figures computed for this article from the published rates above and cross-checked against the ToolMintX API cost calculator