Gemini 4 Argon vs GPT-6.1 Sol: API pricing, discount, and long prompts
Sol costs less than Argon on the half-cached example request. Google lists Argon's discount through November 30, 2026, so check the standard rates before a budget runs past that date. Sol's rates change above 272,000 input tokens.
GPT-6.1 Sol lists 50% less than Gemini 4 Argon for input and 50% less for output at standard rates. That's only the starting line. Argon also carries an introductory discount that Google lists through November 30, 2026, so its bill changes the day the discount ends. Sol adds a different wrinkle. Once a prompt passes 272,000 tokens, the whole request moves to a higher tier. This page works through both effects with real request sizes, so you can see which one matters for your traffic.
Compare base rates, Argon's introductory discount, caching, and Sol's long prompt tier.
By TechCompare · Updated
Google: Gemini 4 Argon
Google AI Studio's own price
- Input
- $4
- Output
- $20
- Cached
- $0.20
per 1M tokens
Price checked 2026-11-07
OpenAI: GPT-6.1 Sol
OpenAI's own price
- Input
- $2
- Output
- $10
- Cached
- $0.10
per 1M tokens
Price checked 2026-11-07
Cost Comparison
Based on 100,000 input tokens (50% cached), 5,000 output tokens, and 100 requests.Prices come live from OpenRouter at each maker's own price where it's listed. Temporary deals are left out.
Side-by-side specs
| Spec | Gemini 4 Argon | GPT-6.1 Sol |
|---|---|---|
| Standard input per 1M tokens | $4.00 | $2.00 (better on this spec) |
| Standard output per 1M tokens | $20.00 | $10.00 (better on this spec) |
| Cache reads per 1M tokens | $0.20 | $0.10 (better on this spec) |
| 100K input + 5K output, no cache hits | $0.50 | $0.25 (better on this spec) |
| 100K input + 5K output, half the input cached | $0.31 | $0.155 (better on this spec) |
| 300K input + 5K output, no cache hits | $1.30 | $1.275 (better on this spec) |
| Introductory discount, input / output per 1M | $2.00 / $10.00 through November 30, 2026 | None running |
| Long prompt rates, input / output per 1M | No separate tier | $4.00 / $15.00 above 272,000 input |
| Batch input / output per 1M | $2.00 / $10.00 | $1.00 / $5.00 |
| Context window | 1,048,576 tokens | 1,050,000 tokens |
| Search grounding | 5,000 prompts free per month, then $14 per 1,000 | Check OpenAI's pricing page |
How they differ
Standard input costs $4 per million tokens on Argon and $2 on Sol. Output costs $20 and $10, and cache reads cost $0.20 and $0.10. A request with 100,000 input tokens and 5,000 output tokens costs $0.50 on Argon and $0.25 on Sol when nothing is cached. With half the input read from cache, it costs $0.31 and $0.155. While Google's introductory discount runs through November 30, 2026, Argon bills $2 for input, $10 for output and $0.10 for cache reads, so the half-cached request costs $0.155 instead of $0.31. Sol costs no more than Argon on this request, even with the discount applied. Google's page says standard rates return on December 1, and that the cache is implicit only from then on. A budget that runs past November needs the standard figures above, not the discounted ones. Sol changes tiers above 272,000 input tokens and applies $4 for input, $0.20 for cache reads and $15 for output to the whole request. A 300,000 token prompt with 5,000 output tokens and no cache hits costs $1.30 on Argon and $1.275 on Sol. Argon has no separate long prompt tier, so its rates stay flat across the 1,048,576 token window. These examples leave out cache writes, tool fees and retries. The embedded calculator uses live OpenRouter rates, which can depend on provider routing.
Verdict
Run both models on a fixed set of your own tasks and compare billed tokens per accepted result. On price alone, Sol has the edge at every cache share, so Argon has to save enough retries or review time to close that gap. Don't build a long term budget on Argon's introductory figures. Rerun the comparison at its standard rates, since those are what your bill follows after November 30, 2026. For prompts near 272,000 tokens, test Sol at its long prompt rates as well.
Which should you pick?
Choose Gemini 4 Argon
Pick Argon when your work leans on what Google built it for, which is software engineering, math and science problems, and autonomous multi-step agents. It also accepts audio, file, image, text and video input, so a pipeline can send recordings or clips directly. While the introductory discount runs, a short project can finish at lower rates, but plan the budget from the standard figures. Search grounding is the other draw. Google gives 5,000 prompts free each month, shared across Gemini 3 and later models, and then charges $14 per 1,000. There's no free tier for the model itself.
Google's Gemini pricing pageChoose GPT-6.1 Sol
Pick Sol when your agents and coding tools already run on OpenAI's API, since switching providers costs integration and evaluation work that a small price gap rarely repays. Keep an eye on prompt length, because requests above 272,000 input tokens bill at the long prompt tier. Confirm that your input types are supported before you move a workload, and price any tools you rely on from OpenAI's own pricing page, since this page only covers token rates.